Demographic characteristics of the patients
Among 1483 patients with FIGO IVB cervical cancer diagnosed between 2010 and 2016, 336 (22.6%) of them underwent surgery, while 1147 (77.3%) did not. The clinicopathological characteristics of patients with metastatic cervical cancer are summarized in Table1. The median age for the non-surgery group was 56 years, while it was 52 years for surgery group, indicating that older patients were more prone to choose conservative treatment. Race did not appear to affect the decision to excision the primary site tumor. The tumors resected from patients who had cervix surgery were less likely to be squamous cell carcinoma, and more likely to be T1 and T2 (both p<0.0001). Compared to non-surgery group, the surgery group had more patients with negative lymph node (27.4% vs. 36.6%, p=0.001). Less patients with multiple metastatic sites would opt for surgery, and more patients in the surgery group had other site or distant LN affected. There were more patients in surgery group deciding to remove distant metastases and received systemic chemotherapy. More patients received radiotherapy if they did not undergo cervix surgery.
Analysis of CCSS and OS
To analyze the effects of local surgery on CCSS and OS in patients with metastatic cervical cancer, log-rank test were performed between surgery and non-surgery groups. As illustrated in Fig. 1, locoregional surgical treatment conferred a prominent survival advantage. The median CCSS and OS were 37 and 28 months for the surgery group, 13 and 11 months for the non-surgery group (Fig. 1A and Additional file 1).
In order to eliminate the bias of site-specific metastasis on survival analysis, data were stratified based on the distant organs involved (multiple site, bone, liver, brain and lung, other site or distant LN), and CCSS was evaluated based on whether or not surgery of the primary tumor was performed in mentioned-above patient groups. Owing to the limited number of patients with isolated brain metastasis, they were excluded from this evaluation. Patients with isolated lung and other site or distant LN involved benefitted from primary site surgery (p = 0.03 for lung only; p < 0.0001 for other site or distant LN) (Fig. 1E, F). Nevertheless, the locoregional treatment could not prolong the survival of patients with bone, liver and multiple sites metastasis. Furthermore, the aggregate effects of surgical resection of primary and metastatic lesions on survival were also evaluated. It showed that survival was better for patients who underwent resection for distant organs, and significantly better for patients whose primary cervical tumors were resected (Fig. 2A).
Local surgeries were categorized into the “Hysterectomy” group (without regional lymph node dissection) and “Radical hysterectomy” group (with regional lymph node dissection). In all cases, pelvic lymph node dissection provided additional survival advantage to the cervical cancer patients (Fig. 2B).
The effects of chemotherapy and/or radiotherapy together with local surgery on survival in patients with metastatic cervical cancer were next analyzed (Fig. 2C-F). Prognosis was better for patients who underwent locoregional resection combined with chemotherapy or radiation compared with those received surgery/chemotherapy/radiation alone. Chemotherapy conferred similar OS and CSSS to locoregional surgery, while radiation provided worse survival compared to surgical treatment.
Data stratified on the basis of histology, and lymph node status showed that they did not affect the survival advantage provided by locoregional surgery. More advanced AJCC T stage, less advantage gained from primary site surgery. Eventually, survival benefit disappeared for T4 tumor.
Univariate and multivariate analysis
Univariate and multivariate analysis were then performed to evaluate the effects of the clinicopathological factors on CSSS and OS (Table 2). As illustrated in Table 2, patients with age between 41 and 80 were at a lower risk of cancer-related death compared to those younger than 40. Compared to white race, black women were at a higher risk of mortality. Tumors with higher AJCC T stage had a poor prognosis. When the lymph node was affected, CCSS was worse than in cases with negative nodes. For treatment, regional LN surgery, chemotherapy and radiation prolonged cancer-related and overall survival for cervical cancer with primary metastasis, while surgery of distant metastasis did not show any significance in CSSS and OS. Except for only brain metastasis, women with isolated organ metastases had better OS and CSSS compared to those with multiple distant organ involved.
Local cervix surgery reduced the cancer-associated and overall mortality rate by about 33% and 31%, which demonstrated that regional resection of primary tumor is an independent prognosis factor and is of vital significance in prolonging the survival of cervical cancer patients with primary metastasis.