Clinical Significance Of Micrometastasis Detection In Lymph Nodes Of Patients With Early-Stage Cervical Cancer

Objects 1. To investigate the risk of postoperative regional recurrence and distant metastasis in patients with early-stage cervical cancer undergoing radical hysterectomy and to analyse the adverse factors which can influence the therapeutic effects of operation and disease-free survival (DFS). Also the significance of detection of lymph node micrometastasis (LNmM) and the necessity of actively adjuvant therapy in early-stage cervical cancer were demonstrated. 2. To approach the feasibility and reliability of detection of LNmM by immunohistochemical analysis in early-stage cervical cancer, which were revealed negative with conventional pathologic techniques, to determine the clinical and pathological factors related to LNmM, and to assess the clinical utility of micrometastasis.Methods 1. We retrospectively reviewed the medical records of 233 patients who underwent radical hysterectomy and pelvic lymphadenectomy for FIGO stage I~II cervical cancer at Shandong University Qilu Hospital from Jan, 1997 to Mar, 2006. The median age was 47.7 years (range 22-72 years), and 106 cases (45.5%) were stage I and 127 (54.5%) were stage II. The series of patients consisted of 188 cases (80.7%) of squamous cell carcinoma, 36 cases(15.4%) of adenocarcinoma and 9 cases (3.9%) of adenosquamous carcinoma. Of all the 233 cancers, 18(7.7%) were well differentiated, 116(49.8%) were moderately differentiated and 99(42.5%) were poorly differentiated. The median follow-up was 69 months (range 11-118 months). Disease-free survival, overall survival (OS), rate of postoperative regional recurrence and that of postoperative distant metastasis were calculated. Relative factors such as patients age, clinical stage, tumor size, tumor growth pattern, histological type, tumor differentiation, depth of stromal invasion, parametrial invasion, pelvic lymph node metastasis, corpus uteri involved, vaginal margin involved, lymphovascular space invasion (LVSI) and ajuvant therapy which may be correlated with postoperative DFS, regional recurrence and distant metastasis in early-stage cervical cancer were analyzed. The frequency of lymph node metastasis were counted and the DFS and OS of cases without lymph node metastasis, with only 1 lymph node involved and with 2 or more lymph nodes involved were calculated, respectively. 2. Medical records were collected and reviewed retrospectively of 67 patients, who were primarily diagnosed as stage I~II cervical cancer, underwent radical hysterectomy and pelvic lymphadenectomy in Shandong University, Qilu Hospital from Jan, 1999 to Dec, 2002, and were histologically negative with conventional pathologic techniques for lymph node metastasis. The median age was 47.3 years (range 22-70 years). Thirty two patients had stage I disease (47.8%) and 35 had stage II disease (52.2%). Of the histological subtypes, fifty-five (82.1%) were squamous cell carcinomas, nine (13.4%) were adenocarcinomas, and 3(4.5%) were adenosquamous carcinomas. Of all the 67 cases, 8 (12.0%) were well- differentiated, 36 (53.7%) were moderately differentiated and 23(34.3%) were poorly differentiated. The median follow-up was 71 months (range 46-94 months). A total of 649 formalin-fixed paraffin-embedded pelvic lymphadenectomy specimens were serially sectioned and immunohistochemically stained with antibodies of cytokeratin (CK) 19 and CK AE1/AE3, among which 78 were parametrial lymph nodes, 183 were obturator lymph nodes, 201 were iliovascular lymph nodes and 187 were inguinal lymph nodes. The expressive frequency of the two indices was compared. The relationship between lymph node micrometastasis and clinicopathological features, such as patients age, clinical stage, tumor size, tumor growth pattern, histological type, tumor differentiation, depth of stromal invasion, parametrial invasion, corpus uteri involvement, vaginal margin involvement, and LVSI was analyzed. The contribution of micrometastasis to postoperative recurrence/metastasis and survival rate in early cervical cancer was investigated. Results 1. Of all cases, forty-one was registered postoperative regional recurrence, with the recurrent rate 17.6% and the risk of recurrence was significantly related to parametrial invasion, depth of stromal invasion and vaginal margin involved(P<0.05). Distant metastasis were reported in 23 cases, with the rate 9.9%, and parametrial invasion, histological type, pelvic lymph node involvement and LVSI were significant factors associated with metastasis (P<0.05) . The 3- and 5- year DFS of the full set were 84.2% and 63.5% respectively, and the corresponding OS were 86.3% and 69.2%, respectively. Among the variables, clinical stage (II), tumor size (≥4cm), histological type (non-squamous carcinoma), tumor differentiation (poorly differentiation), pelvic lymph node metastases, parametrial invasion, vaginal margin involved, deep stromal invasion and LVSI were the poor prognostic factors in univariate survival analysis (P<0.05). Cox proportional hazard model analysis showed that only pelvic lymph node metastases and LVSI were independent survival predictors. The rate of lymph node metastasis of the full set was 24.9%(58/233). The 5-year DFS of cases without lymph node metastasis, those with only 1 lymph node involved and those with 2 or more lymph nodes involved were 73.8%, 64.6% and 32.2% respectively, and the corresponding overall survival were 80.9%, 67.4% and 33.5% respectively, and the difference was significant (P<0.05) . 2. Of all the 649 negative lymph nodes from 67 patients, 34 lymph nodes from 13 patients were positively stained by anti-CK antibodies immunohistochemical methods, among which 4 were parametrial lymph nodes, 19 were obturator lymph nodes, 7 were iliovascular lymph nodes and 4 were inguinal lymph nodes. Micrometastasis was detected in 13 of 67 (19.4%) patients, found in 34 of the 649 (5.2%) lymph nodes evaluated. The highest positive frequency of 10.3%(19/183) was detected in obturator lymph nodes, which was significantly higher than that of other groups of lymph nodes(P <0.05). Of the 13 cases with micrometastasis, expression of CK19 was detected in 10 patients, expression of CK AE1/AE3 was positive in 12 patients and both two were found in 9 patients. Of the 34 lymph nodes with micrometastasis, expression of CK19 was detected in 26 nodes, expression of CK AE1/AE3 was positive in 31 nodes and both the two were found in 23 ones. There was no significant difference between the detection rate of the two antibodies in patients and lymph nodes(P>0.05). Univariate regression analysis showed that LNmN was correlated with clinical stage, histological type, LVSI and parametrial invasion, among which only LVSI and parametrial invasion were independent risk factors by multivariate analysis. At the end of follow-up, 5 out of 13 patients with micrometastasis had recurrent disease, 2 patients developed distant metastasis and 4 died, while 6 of 54 patients without micrometastasis had recurrent disease, 3 developed distant metastasis and 4 died. There was significant difference between the recurrent/metastatic rates of the two groups(P <0.05). The 3-, 5- year DFS of patients with or without micrometastasis were 76.9%, 68.4% and 92.5%, 83.8% repectively, and the corresponding OS were 84.3%,74.9% and 94.4%,92.4%, respectively. The difference of survival rates between the two groups was significant(P<0.05).Conclusions 1.There are still possibilities for post-operative tumor recurrence or metastases in patients with FIGO stage I～II cervical cancer, and parametrial invasion, depth of stromal invasion and vaginal margin involved are significant factors associated with recurrence, while histological type, parametrial invasion, lymph node involvement and LVSI are important factors related to metastasis. Forty or more years old, Stage II, tumor diameter to be or more than 4cm, non-squamous carcinoma, poor differentiation, pelvic lymph node metastasis, parametrial invasion, vaginal margin involved, deep stromal invasion and LVSI are all suggestive of poor prognosis, while only pelvic lymph node metastasis and LVSI are independent survival predictors. The survival of patients with lymph nodes metastasis decreases dramatically with the increase of the number of lymph nodes involved. 2. Micrometastasis is identifiable in negative lymph nodes detected by conventional pathological technique. With CK19 and CK AE1/AE3 as markers of micrometastasis, immunohistochemistry is able to increase the detection rate of lymph node metastasis in cervical cancer, and there is no significant difference in detection rate between the two markers(P＞0.05). The elevated detection rate in obturator lymph nodes suggests that micrometastasis is likely to involve in a first lymph-node drainage station. Lymph node micrometastasis is more frequently detected in early cervical cancer patients with parametrial invasion and LVSI and LNmM is predictive of lower postoperative survival rate and poor prognosis. Detection of LNmM is of practical significancy for cervical cancer patients in evaluating prognosis and instructing treatment more accurately. Combined therapyof additional radiotherapy and/or chemotherapy after operation may prevent ordecrease the risk of poor prognosis resulting from micrometastasis.