Endometrial cancer is the most common gynecological malignancy in Germany, as well as in the developed world and globally [1, 2, 3]. Approximately 70% of patients with endometrial cancer present at an early stage, when the tumor remains confined to the corpus uteri, and have a very good 5-year-survival rate of over 80% [1, 4]. The surgical staging of endometrial cancer was first recommended by the Fédération Internationale de Gynécologie et d’Obstétrique (FIGO) in 1988 [5], replacing the clinical staging that was earlier practiced and was based exclusively on clinical findings, reflecting a paradigm shift in disease management. Since the introduction of the new FIGO recommendations, discussions and debates have been ongoing about the value of lymphadenectomy not solely as a staging procedure, but as a treatment modality in itself. Randomized controlled trials have found no survival benefit associated with the procedure [6, 7], while some retrospective cohorts [8, 9, 10] as well as a meta-analysis [11] have found the contrary. In a recent Cochrane review, lymphadenectomy was regarded as a means of guiding the need for further treatment, such as irradiation, but not as a therapeutic means in itself [12].

Pelvic irradiation is complementary to the role of surgery in the treatment of endometrial cancer. Two large prospective studies (the PORTEC-trial and the GOG-99 trial) presented evidence that pelvic irradiation provides a benefit in the form of a longer disease-free interval but does not provide any overall survival benefit [13, 14]. Another randomized trial (PORTEC-2) showed that local irradiation using brachytherapy could be as effective as pelvic irradiation in reducing pelvic recurrences but had a significantly lower morbidity [15].

The Surveillance, Epidemiology and End Results program (SEER) includes data from 17 different cancer registries in the United States. The aim of this study was to evaluate the therapeutic role of lymphadenectomy and irradiation in the treatment algorithm of high-risk endometrioid endometrial cancer.

This was a population-based retrospective cohort analysis using the freely accessible public SEER database. SEER data from patients with endometrial cancer treated between 2004 and 2012 were accessed online on March 1, 2016 after obtaining the required authorization (SEER ID: 14859-Nov2015). Program version 8.2.1 was used, which included data collected prior to the November 2014 submission.

Patients were selected from the SEER database records from the period 2004-2012 who were registered with the diagnosis of endometrial endometrioid cancer. Patients who had incomplete records such as unknown or missing regional lymph node status, those with lymph node status examined using techniques other than lymphadenectomy, those with tumor status unknown, not examined or not found were excluded. Patients with cause of death other than cancer were also excluded. Between 2004 and 2012, there were 47,463 registered patients with endometrial cancer of the endometrioid type, of whom only 41,612 were available for statistical analysis after excluding patients not meeting the inclusion criteria (Fig. 1).

Fig. 1.

Algorithm showing the process of patient selection for the statistical analysis from the SEER data set. The total number of patients included as well as those who were excluded according to the inclusion and exclusion criteria are thoroughly depicted.

The mean age in the patient dataset $\pm$ standard deviation (SD) was 60.14 $\pm$ 11.36 years. A total of 10,288 patients were of high risk, 27,411 were of low risk, and 3,913 were of advanced stage. Mean survival $\pm$ SD was 45.5 $\pm$ 30.61 months, ranging from 0-107 months. The demographic data, stage, lymph node status, and irradiation status of the patients who met the inclusion criteria are shown in Table 1.

Among the 26,339 (63.3%) patients who underwent lymphadenectomy, 2,427 (9.2%) had positive lymph nodes, whereas in the high-risk group 1,080 (13.2%) had positive lymph nodes. In the high-risk group, 8,292 (80.6%) patients underwent lymphadenectomy whereas 1,996 (19.4%) patients did not receive an operative examination of the lymph node status. Among those, 1,052 did not receive any sort of adjuvant treatment while 944 received irradiation. The survival outcome of the high-risk patients who were divided into six groups were compared pairwise and the results are shown in Table 2. Kaplan-Meier survival curves for the six subgroups, in addition to the low-risk and advanced disease groups, are shown in Fig. 2. Patients in the high-risk group who received a lymphadenectomy that showed lymph node involvement but did not receive pelvic irradiation had the worst overall survival (median overall survival of 80 months) among the entire high-risk group. This group’s survival rate was comparable to that of patients in the advanced disease group.

Fig. 2.

Kaplan-Meier curves showing the overall survival in the patients with low-risk, high-risk and advanced disease. The high-risk patients were divided into six groups: H1: no Lymphadenectomy & no irradiation, H2: no lymphadenectomy & irradiation, H3: positive lymph nodes & no irradiation, H4: positive lymph nodes & irradiation, H5: negative lymph nodes & no irradiation, H6: negative lymph nodes & irradiation. Patients in the high-risk group who received a lymphadenectomy that showed lymph node involvement but did not receive pelvic irradiation (group H3) had the worst overall survival.

Table 3 shows the survival outcomes of patients subdivided according to lymph node status and percentage of involved lymph nodes out of the total number of lymph nodes removed. In keeping with the expected prognostic value of surgical staging overall survival was inversely proportional to the percentage of lymph nodes involved in a statistically significant manner. A similar outcome was observed for the high-risk group of patients yielding an. Please see Table 4 for full statistical details of all patient groups.

Table 5a and Table 5b show the survival outcomes of patients according to the total count of the lymph nodes removed in both the whole cohort and in the high-risk group of patients, respectively. Among the whole patient cohort, a consistent significant rise in overall survival was observed with increasing count of the lymph node yield until a count of 40 lymph nodes was reached. A lymph node yield greater than 40 did not confer any further significant overall survival benefit. A similar pattern was also observed in the high-risk group of patients. Fig. 3 (a) shows Kaplan-Meier Curves of the overall survival according to the number of lymph nodes removed in the entire cohort and (b) in the high-risk group of patients. Table 6 and Table 7 show the survival outcomes of patients according to the absolute number of infiltrated lymph nodes in the patient cohort as a whole and in the high-risk subgroup of patients, respectively. A significant worsening in overall survival was observed with increasing infiltrated lymph node count, both within the patient cohort as a whole and within the high-risk sub-group.

Fig. 3.

(a). Kaplan-Meier curves showing the overall survival in the patients who received a lymphadenectomy according to the number of removed lymph nodes and (b). Kaplan-Meier curves showing the overall survival in the high-risk patients who received a lymphadenectomy according to the number of removed lymph nodes. A lymph node yield greater than 40 did not confer any further significant overall survival benefit.

The survival outcome for the high-risk group, with respect to lymph node status and type of radiation received, if any, revealed a statistically significant effect on survival in the group of patients who received external beam irradiation after lymphadenectomy in comparison with those who did not receive any type of radiation (Table 7a, Table 7b and Table 7c).

Since the introduction of systematic lymphadenectomy to the operative staging process for endometrial cancer in 1988 [5], the therapeutic value of this procedure has been highly debated. Prospective randomized trials have shown no survival benefit for lymphadenectomy [6, 7]. However, these trials have been heavily criticized for being underpowered, for not having clearly defined the role of lymphadenectomy in guiding therapy, and for the suboptimal quality of the lymphadenectomy procedure used [16, 17, 18].

Our data analysis produced some apparently contradictory results. On the one hand, a mean overall survival of 80 months was demonstrated in the high-risk group of patients with lymph node involvement who had not received any form of adjuvant treatment. This group of patients showed the worst prognosis among the high-risk collective, despite the average overall survival of 80 months, reflecting some therapeutic benefit associated with the procedure or, perhaps, could be explained by the dormant biological nature of lymph node metastasis. On the other hand, while analyzing the survival outcome in the high-risk group with respect to the percentage of lymph node involvement to the total number of extracted lymph nodes, we demonstrated a significant linearly progressive worsening in survival with increasing percentage of involved lymph nodes. This finding is also consistent with the results of two prospective studies [6, 7] as well as a recent retrospective study [19]. We consider this finding to negate the therapeutic value of lymphadenectomy. If the procedure had provided a survival benefit, it would have been manifested in observing a better or at least comparable survival rate between groups regardless of the percentage of lymph node involvement.

As regards the quality or efficiency of the lymphadenectomy procedure itself, our results showed a positive correlation between survival and increasing lymph node yield until a plateau of 40 or more lymph nodes was reached, above which the extraction of additional lymph nodes did not provide any further survival benefit to the patient. Chan et al. (2006) have also reported a limited survival benefit associated with the extraction up to a maximum threshold number of lymph nodes; in their study the threshold above which no further survival benefit of lymph node removal in high risk endometrial cancer patients occurred at 20 lymph nodes [20]. Although the threshold observed in our study’s patients was twice as high as that reported in the Chan et al., study the overall finding of a limited survival benefit for lymphadenectomy is consistent between them.

Lymphadenectomy in early-stage high-risk endometrial cancer is considered an important tool in guiding further adjuvant treatment modalities. Prospective randomized trials have shown no survival benefit for radiotherapy in intermediate/high-risk endometrial cancer patients [13, 14], but have demonstrated a significant progression-free interval that could be attained by brachytherapy alone without external pelvic irradiation [15]. Our results, derived from a US population-based database, are concordant with these findings. Our results indicate that administering radiotherapy to high-risk patients without determining lymph node status did not provide any survival benefit, supporting the current role of lymphadenectomy in the diagnostic algorithm of endometrial cancer. In fact, in this study’s cohort, administering external beam irradiation without having determined lymph node status by lymphadenectomy was associated with a marginally worse overall survival.

Our study’s findings reflect the implications of treatment recommendations on a US based population. Many patients with histologically defined disease-free lymph nodes received pelvic irradiation, while other high-risk patients that should have received lymphadenectomy did not undergo this procedure but were nonetheless treated with pelvic irradiation that marginally worsened the outcome. We were unfortunately unable to extract the basis for such decisions from the data.

Limitations of this study include the retrospective nature of the analysis, missing information regarding standardization of indications as well as likely variation in the performance efficiency of the lymphadenectomy procedure itself. Furthermore, the extent of regional lymph node resection was not clearly defined in the source data, meaning that we could not conclude whether para-aortal lymph node dissection was performed in all cases, as recommended [21, 22], and as shown in a recent retrospective analysis to provide a survival benefit in high-risk endometrial cancer patients [23]. Another weakness of this study is that there was no information recorded regarding the type of radiation therapy performed and this is likely to have varied.

From this analysis, we could not provide clear evidence that lymphadenectomy confers a survival benefit to patients with endometrioid endometrial cancer. However, the current role of lymphadenectomy in staging and as a tool for guiding further adjuvant treatment was supported. Well-designed prospective randomized trials are required to determine the therapeutic value of lymphadenectomy in patients with high-risk endometrioid endometrial cancers.

AGO, German workgroup on gynecologic oncology (Arbeitsgemeinschaft Gynäkologische Onkologie); CI, confidence interval; FIGO, Fédération Internationale de Gynécologie et d’Obstétrique; OS, overall survival; SD, standard deviation; SEER, Surveillance, Epidemiology and End Results program.

All authors contributed to study design and idea formulation. AAS, MSS, SAM, SA, AN and BR were responsible for data acquisition and tabulation. AAS, MSS, SAM, OG and EM contributed to statistical analysis and data editing. All authors contributed to manuscript writing, revision and finalizing in its current form, and are hence responsible for the work in its current form.

This article does not contain any studies with human participants performed by any of the authors.

Thanks to all the peer reviewers and editors for their opinions and suggestions.

The authors declare no competing interests.