Multivariate survival analysis of the patients with recurrent endometrial cancer

The number of patients with endometrial cancer is increasing in Japan as well as in the United States and other countries [1]. A majority of endometrial cancer patients are diagnosed as being without clinical evidence of extrauterine spread (FIGO stage I and II) and have a 5-year survival of approximately 90%. Approximately 10% to 15% of patients with early stage endometrial cancer will experience recurrences [2, 3]. To reduce the recurrence rate, adjuvant chemotherapy or radiotherapy has been applied. However, a definite standard therapy has not yet been established.

Recently, the number of patients with recurrent endometrial cancer and advanced-stage disease is also increasing. For stage III-IV endometrial cancer, Randall et al. [4] recently reported that platinum-doxorubicin (AP) chemotherapy showed significantly better survival effect than whole abdominal irradiation (WAI) as an adjuvant treatment for advanced-stage endometrial cancer. Although cytotoxic chemotherapy has been used primarily as palliative therapy for patients with advanced endometrial cancer, there is relatively little experience with postoperative systemic chemotherapy used as an adjuvant treatment in western countries. However, with the introduction of routine surgical staging, we have performed extensive surgery including systematic pelvic and para-aortic lymphadenectomy and post-surgical cisplatin-based systemic chemotherapy for patients with intermediate or high risk of recurrence [5, 6]. We have recently reported the prognosis and failure pattern of endometrial cancer patients treated with our treatment strategy, demonstrating better survival of node-positive patients with endometrial cancer compared to that reported in the literature [6]. We also revealed that distant failure was the most prevalent pattern of failure in spite of systemic adjuvant chemotherapy, suggesting that we need to optimize primary treatment and treatment after relapse to further improve survival of endometrial cancer [6].

While endometrial cancer is a highly curable malignancy when it presents as a uterine-confined disease, the prognosis for recurrent or metastatic disease is poor. The median survival of women enrolled in trials for recurrent or metastatic endometrial cancer hardly exceeds 12 months [7]. There is still no agreement in the literature regarding the most adequate treatment for those patients. Thus, it is important to establish a prognostic model to predict survival after relapse in endometrial cancer.

Therefore, the aim of this retrospective study was to determine the factors influencing survival of relapsing patients, and to assess an alternative strategy to routine adjuvant chemotherapy.

In the present study, we analyzed the prognosticators for survival after relapse in patients with recurrent endometrial cancer who underwent primary complete cytoreductive surgery combined mainly with adjuvant chemotherapy. We first demonstrated that time to relapse after primary surgery and the number of relapse sites are independent prognostic factors for survival of recurrent endometrial cancer in this study.

In the literature, prognostic analyses have been reported for patients with recurrent endometrial cancer who received primary surgery and adjuvant radiotherapy. Previously reported studies had indicated histological grading of the tumor [9, 10] and time to progression interval [10] as the most important prognostic factors in the recurrent and metastatic cancer setting. Sartori et al. [11] previously reported that site of relapse, the interval of time from original surgery to recurrence, and whether postoperative pelvic radiotherapy was administered were independent prognostic variables for survival by multivariate analysis in recurrent endometrial cancer. Sohaib et al. [12] reported that multivariate analysis identified multiple sites of disease, liver and splenic metastases to be independent predictors of poor outcome in recurrent endometrial cancer. Thus, this study may be the first report, as far as we know, on the prognostic analysis for patients with recurrent endometrial cancer treated with primary surgery and adjuvant chemotherapy.

Although the exact reason why two factors are independently related to survival after relapse remains to be determined, it is easy to speculate the possible reasons. Time to relapse within one year after primary surgery may largely depend on the aggressive phenotype and resistance to adjuvant chemotherapy of endometrial cancer cells, and multiple sites of recurrence may be related to the metastatic potential of endometrial cancer cells. Therefore, we need to establish a new treatment strategy to efficiently treat patients with recurrent endometrial cancer in addition to currently available treatment modalities including debulking surgery, chemotherapy, and radiotherapy. One of the promising molecular targeting agents is the vascular endothelial growth factor inhibitor, bevacizumab, which has been recently shown to improve progression-free survival of the patients with ovarian cancer [13]. To search for new molecular markers to predict survival after relapse and/or new molecular targets to improve survival, we need to further investigate the molecular mechanism of chemoresistance and radioresistance in endometrial cancer cells. Cyr61, a member of CCN family, may be a good target because Cyr61 predicts survival of patients with endometrial cancer [14], and has been reported to induce resistance to carboplatin in ovarian cancer cells [15] and resistance to paclitaxel in breast cancer [16]. Additionally, we need to investigate the molecular mechanism on the metastasis and invasion process of endometrial cancer cells.

Complete resection of recurrent tumors has been reported to be a prognosticator for recurrent endometrial cancer in the literature [17], while this is not related to survival in our patient cohort. This may be due to the small number of analyzed cases. Another possibility is because debulking surgery is usually applied to selected patients with a single relapse site, which was independent prognostic factor for survival after relapse in this study. Histologic type was significantly related to survival after relapse in univariate analysis, whereas not after multivariate analysis in this study. This is probably because the patients with non-endometrioid histology tend to recur with shorter progression-free survival than those with endometrioid types, and time to relapse within one year after primary surgery was an independent prognostic factor in this study. Similar to all retrospective studies, the number of patients is fairly small and our findings need to be further validated in larger cohorts.

Since the prognosis of the patients with recurrent endometrial cancer generally remains poor, and risk for recurrence showed marginal significance for survival after relapse in this study as well as in our previous study [17], we need to reconsider the current best treatment strategy as a primary treatment. We have recently shown that para-aortic lymphadenectomy significantly improves recurrence-free survival and overall survival of endometrial cancer patients with intermediate or high risk for recurrence [18]. Therefore, we should offer a better treatment strategy consisting of extensive surgery including systematic para-aortic lymphadenectomy and systemic adjuvant chemotherapy to provide better survival benefit for patients with intermediate/high risk for recurrence. On the contrary, endometrial cancer patients with low risk for recurrence may not need to receive neither adjuvant treatment nor extensive lymphadenectomy including para-aortic nodes as previously reported by Mariani, et al. [19].

In summary, time to relapse after primary surgery and the number of relapse sites are independent prognosticators for survival after relapse in patients with recurrent endometrial cancer. We need to offer the opportunity to participate in clinical trials to establish new treatment strategies using molecular targeting drugs for recurrent endometrial cancer patients. We also need to further investigate the molecular mechanisms governing sensitivity to chemotherapy and radiotherapy, and the metastatic processes in endometrial cancer cells to establish new treatment strategy for recurrent cases.