Impact of ovarian metastatectomy on survival outcome of colorectal cancer patients with ovarian metastasis: a retrospective study

2.1 Patient characteristics

Patients’ clinical characteristics were summarized in Table 1. In total, 68 patients newly diagnosed with ovarian metastatic CRC were identified from a cohort of 2170 female patients with metastatic CRC, including 54.4% premenopausal patients. Hence, the incidence of ovarian metastases in female metastatic CRC was 3.1%. The median age at diagnosis of ovarian metastasis was 46.5 years old. The median time from primary cancer diagnosis to ovarian metastasis detection was 4.5 months. And the median largest diameter of ovarian lesions was 8.0 cm. There were 32 (47.1%) cases presented with synchronous ovarian metastasis, while 36 (52.9%) with metachronous metastasis. Twelve (17.6%) patients had metastases confined in the ovaries, whereas the left 56 (82.4%) patients presented with extraovarian metastases in liver, lung, peritoneum, bone, and et al. when diagnosed with ovarian metastatic CRC. More than half of the patients (52.9%) had ascites and 41.2% had peritoneal metastasis. As to treatment approach, 42.6% of all cases underwent complete resection of ovarian metastases, 41.2% cases received palliative resection, and the left 16.2% patients did not undergo ovarian metastatectomy. Most patients (70.6%) received systemic chemotherapy after diagnosis of ovarian metastasis. In the group of no resection, all the patients received systemic chemotherapy, while in the groups of complete and palliative resection, 55.2% and 75% patients received chemotherapy, respectively.

2.2 Survival analyses for different groups

In total, the median DSS after diagnosed with ovarian metastasis was 25.0 months (95% confidence interval (CI) = 21.0–29.0 months). Univariate survival analysis results were listed in Table 1. In this study, we classified surgical methods of ovarian metastases into three groups, i.e. no resection (chemotherapy only), palliative resection (R1 and R2 resection) and complete resection (R0 resection). Applying Kaplan-Meier method, we found that ‘complete resection group’ showed significantly longer DSS than that of the other two groups (‘complete resection group’ vs. ‘palliative resection group’: 33.0 vs. 20.0 months, P = 0.000; ‘complete resection group’ vs. ‘no resection group’: 33.0 vs. 22.0 months, P = 0.026, Fig. 1), while median DSS of ‘palliative resection group’ was similar to that of ‘no resection group’ (20.0 vs. 22.0 months, P = 0.630, Fig. 1). These results suggested that complete resection of metastatic ovaries predict better survival in CRC patients.

Stratified analysis revealed that in patients without systemic chemotherapy, the median DSS of complete resection group was significantly longer than that of palliative resection group (26.0 vs. 11.0 months, P = 0.001, Table 2, Fig. 2A). In patients receiving systemic chemotherapy, complete resection also prolonged patients’ survival time (vs. palliative resection: 69.0 vs. 21.0 months, P = 0.002; vs. no resection: 69.0 vs. 22.0 months, P = 0.007), while palliative resection could not improve survival outcome compared with no resection (21.0 vs. 22.0 months, P = 0.993, Table 2, Fig. 2B).

2.3 Prognostic factors associated with survival

Using multivariate Cox proportional hazard regression analysis, we determined the associations between survival and multiple variables such as age, menopausal status, the levels of CEA, primary tumor location, T and N stage of primary tumor, histological type, differentiation grade, primary tumor resection, ascites, synchronous metastasis, bilateral metastasis, size of ovarian metastases, peritoneal metastasis, liver metastasis, lung metastasis, surgical methods of ovarian metastases and systemic chemotherapy, et al. The results demonstrated that regional lymph nodes metastasis of primary tumor (hazard ratio (HR): 3.438, 95%CI: 1.094–10.810, P = 0.035), primary tumor resection (HR: 6.436, 95%CI: 1.770-23.399, P = 0.005), differentiation grade (HR: 0.272, 95%CI: 0.107–0.693, P = 0.006), complete resection of ovarian metastases (versus palliative resection: HR: 17.091, 95%CI: 3.040-96.099, P = 0.001; versus no resection: HR: 9.519, 95%CI: 1.581–57.320, P = 0.014), and systemic chemotherapy (HR: 3.059, 95%CI: 1.089–8.595 P = 0.034) were independent factors for predicting survival outcome (Table 3).

5.1 Patient selection

Female patients who were newly diagnosed with ovarian metastatic CRC between January 2008 and December 2017 were identified retrospectively from the database of Cancer Center at West China Hospital, Sichuan University. Patients’ information and treatment history were obtained through medical records. All patients were followed-up to July 31, 2018.

Data of the patients were collected including age at the time of diagnosis with ovarian metastasis, menopausal status, pretreatment serum carcinoembryonic antigen (CEA) level, primary tumor location, initial T and N stage of primary tumor, time interval between diagnosis of primary tumor and that of ovarian metastasis (metachronous vs. synchronous), size of ovarian metastases, combined metastases outside ovaries, ascites, surgical methods of ovarian metastases (no resection vs. palliative resection vs. complete resection), systemic chemotherapy and response evaluation. For patients received surgical resection of ovarian masses, pathological diagnoses were confirmed by immunohistochemistry staining of ovarian specimens (CK7-/ CK20+/ CDX2+ /CA125-/ PAX8-). In current study, the time interval between diagnosis of primary tumor and that of ovarian metastasis exceeding 3 months was defined as metachronous ovarian metastasis, otherwise as synchronous metastasis. Palliative resection of ovarian metastases meant microscopic or macroscopic residual disease in ovaries (R1 or R2 resection), while complete resection referred to no residual disease left in ovaries (R0 resection).

This study was approved by the West China Hospital, Sichuan University Ethic Committee for Clinical Investigation. And all methods were performed in accordance with the relevant guidelines and regulations.

5.2 Statistical analysis

Median disease specific survival (DSS) was estimated with Kaplan-Meier method, which was calculated from the date of diagnosis with ovarian metastasis to death (caused by cancer) or last follow-up time. Death caused by CRC was considered as endpoint event. Patients who were still alive at last follow-up time or who had died of other causes were censored. Log-rank test was used to compare survival time of different groups and a multivariate Cox proportional hazard regression model was established to determine the associations between survival time and other factors such as age, menopausal status, primary tumor location, ascites, surgical methods of metastases and systemic chemotherapy, et al. Statistical tests were two sided and P < 0.05 was considered statistically significant. SPSS Statistics 22.0 (IBM, Armonk, NY, USA) was applied to perform all the statistical analyses.