| BackgroundCervical cancer is the second common malignancies in both of incidence and mortality rates in worldwide. There are about 500,000 new cases per year in the world, accounting for 5% of all cancers, and approximately 80% of those are in the developing countries.The spread of cervical cancer occurs primarily by means of direct extension and lymphatic metastasis. Cervical intraepithelial neoplasia (CIN) can subsequently progress to invasive cancer. Although this process can be quite slow, once tumor breaks through the basement membrane, extension of the tumor will be quicken up. The rate of pelvic lymph node metastasis would increase rapidly, especially in the presence of deep stroma invasion and lymphovascular embolism. The management decision for cervical cancer mostly depends on the status of tumor invasion and metastasis. The primary therapies for cervical cancer are surgery, radiotherapy or chemoradiotherapy. Surgery and radiotherapy are equally successful treatment options for early stage invasive cervical cancer. Therefore, accurate preoperative assessment of pelvic lymph node metastasis, the depth ofstroma invasion and the status of lymphovascular embolization, is important to choose individual treatment plan, to improve survival rate and reduce complication. At present, computered tomography (CT) and magnetic resonance imaging (MRI) are primary methods for preoperative evaluation of tumor invasion and metastasis, but it is still a difficult clinical problem to predict deep stroma invasion and lymph node metastasis in early stage cervical cancer by MRI or CT.Angiogenesis theory of tumor, firstly described by Judah Folkman in 1969, suggested that angiogenesis be pivital to the uncontrolled proliferation, invasion, and metastasis of tumor. Angiogenesis has been proved to be involved in cell proliferation, invasion and metastasis of the early stage cervical cancer. Therefore, the assessment of tumor angiogenesis may be a novel means of predicting deep stroma invasion, lymphovascular embolization and pelvic lymph node metastasis.Detection of microvessel density immunohistochemically is regarded as a golden standard procedure to quantitate tumor angiogenesis. However, MVD is always assessed in vitro and retrospectively, and preoperative MVD detection is difficult. Preoperative assessment of tumor angiogenesis would be of value clinically for evaluating severity and progression of tumor, as well as for making a proper treatment decision. Tumor vasculature can be examined in vivo by color Doppler ultrasound (CDUS), which could investigate the characters of blood dynamic situation of tumor on the one hand, and quantify tumor vascularity on the other hand.However, CDUS is relative angle and velocity dependent, Heish etal reported that only 46.2% of the early stage cervical cancer showed detectable blood flow signals by color Doppler. With the advantage of relative angle and velocity independence, extended dynamic range and higher sensitivity;several independent studies have demonstrated that power Doppler ultrasound (PDUS) is more sensitive than CDUS in the depiction of tumor vascularity and flow.Thus far, few studies have concerned on the use of PDUS in predicting deep stroma invasion, lymphovascular embolization and pelvic lymph node metastasis of tumor. The aim of this study was therefore to evaluate the accuracy of PDUS parameters in assessment tumor angiogenesis in vivo, and analysis the relationship between PDUS parameters and clinicopathologic characteristics to evaluate the clinical value of these parameters in the early stage cervical cancer.Materials and methodsFrom May 2004 to January 2006, 72 cervical cancer patients with FIGO stage IB-IIA were enrolled . An unhomogeneous mass was found in the cervix of each patient by Transvaginal ultrasound (TVS). All patients underwent radical abdominal hysterectomy and pelvic lymph node dissection as the initial treatment. Transvaginal power Doppler ultrasound was performed to search for blood flow signals of the tumor before surgery. The resistance index (RI), pulsed index (PI), peak systolic velocity (PSV) and vascularity index (VI) were calculated for analysis. Image—pro ? plus software was used to quantify VI. All the tissue samples were obtained after operation. MVD of the tissue samples was assessed by immunohistochemical method (SP method). All clinicopathologic data were obtained. All the data were statistically managed with SPSS 13.0 for windows .The level of P<0.05 was considered statistically significant. Results1. The correlations between MVD and PDCU parameters:There is a significant positive correlation between MVD and VI(r=0.338,P=0.004), but there is no significant correlations between MVD and RI, PI or PSV.2. The relationship of clinicopathological characteristics and PDUS parameters or MVD2.1 The different FIGO stageRI in FIGO stage IB was significantly higher than that in FIGO stage IIA (t=2.158, P=0.034);VI in FIGO stage IB was significantly lower than that in FIGO stage IIA (t=2.494, P=0.015);MVD in FIGO stage IB was significantly lower than that in FIGO stage IIA (t=2.176,P=0.033);and other parameters had no statistical difference among different FIGO stage.2.2 The different histological type and gradePDUS parameters and MVD had no statistical difference among different histological type and grade.2.3 The different depth of stroma invasionVI in cases with deep stroma invasion was significantly higher than that in cases with superficial stroma invasion (t=4.951,P=0.000);MVD in cases with deep stroma invasion was significantly higher than that in cases with superficial stroma invasion (t=3.164,.P=0.0023);and other parameters had no statistical difference.2.4 With and without lymphovascularity embolismVI in casess with lymphovascular embolism was significantly higher than that in cases without lymphovascular embolism (t=5.299 ,P=0.000);MVD in casess with lymphovascular embolism was significantly higher than that in cases without lymphovascular embolism (t=2.455,P=0.017);and other parameters had no statistical difference2.5 With and without pelvic lymph node metastasisVI in cases with pelvic lymph node metastasis was significantly higher than that in cases without lymph node metastasis (t=5.104, P=0.000);MVD in cases with pelvic lymph node metastasis was significantly higher than that in cases without lymph node metastasis (t=6.309 P=0.000), and other parameters had no statistical difference .2.6 The correlation of tumor size with PDUS parameters and MVDThere were significant positive correlations between tumor size and MVD or VI (tumor size vs MVD r=0.494,P=0.000;tumor size vs VI r=0.414,P=0.000), There was significant negative correlation between tumor size and MVD (r=-0.257 P=0.029), but there was no significant correlations between tumor size and PSV. 3.The significance of VI in predicting deep stroma invasion, lymphovascular embolization and pelvic lymph node metastasis, and determinating the cut off value3.1 ROC curve showed VI values 13% was superior to any other standards in predicting the deep stroma invasion. The sensitivity of this cut off value was 74%, and the specificity was 76%.3.2ROC curve showed VI values 14% was superior to any other standards in predicting lymphovascular emboli. The sensitivity of this cut off value was 94%, and the specificity was 74%3.3ROC curve showed VI values 14% was superior to any other standards in predicting lymph node metastases. The sensitivity of this cut off value was 89%, and the specificity was 78% Conclusions1. VI was significant correlated with MVD. Thus, VI could be used in vivo assessment of tumor angiogenesis2. VI showed more worthful than RI, PI and PSV in predicting deep stroma invasion, lymphovascular embolization and pelvic lymph node metastasis. VI could be used as an important indicator for predicting tumor invasion and pelvic lymphnode metastasis. |
PDF Full Text Request