Lymphadenectomy in women with endometrial cancer: aspiration and reality from a radiation oncologist’s point of view

To investigate the meaning of lymphadenectomy (LNE) in women with endometrial cancer (EC) for clinical outcome and secondly to determine the impact of the method of adjuvant radiotherapy (RT) on survival as well as to define prognostic factors. 322 patients (pts) underwent adjuvant RT for endometrioid EC at our department from 2004 until 2012 and were included in this retrospective study. Chi-square test, LogRank test and Cox regression were used for statistical analyses. Median age at diagnosis: 66 years. FIGO stages: FIGO I 69.4 %, FIGO II 15.3 %, FIGO III 14.5 %, FIGO IV 0.9 %. Surgical staging: 30.6 % pelvic/paraaortic LNE, 45 % sole pelvic LNE, 8.8 % sampling of suspicious lymph nodes, 15.6 % no LNE. Adjuvant chemotherapy (ChT): 3.2 %. Sole intravaginal brachytherapy (IVB): 60.2 %. IVB + external beam radiotherapy (EBRT): 39.8 %. 5-year local recurrence free survival (LRFS): 90.6 %, distant metastases free survival (DMFS): 89.8 %, overall survival (OS):79.3 %. In multivariate analysis age (p = .007), pT stage (p = .029), lymph node status (p = .003), grading (p = .011) and lymphovascular space invasion (LVSI; p = .008) remained as independent prognostic factors for OS. Resection status (p = .01) and LVSI (p = .014) were independent prognostic factors for LRFS and LVSI (p = .008) was the only independent prognostic factor for DMFS. There was no statistically significant survival benefit from LNE in LRFS (p = .561), DMFS (p = .981) or OS (p = .791). 5-year LRFS in stage I and II: 96.0 and 82.9 % after sole IVB, 90.8 and 81.6 % after combined IVB/EBRT (p = .105; p = .970). 5-year OS rates for stage I and II: 86.5 and 71.3 % after sole IVB, 84.2 % and 69.2 % after combined IVB/EBRT (p = .153; p = .619). Comprehensive surgical staging is rarely performed and may be omitted in women with endometrioid EC in stages I-II. Sole IVB delivers equally good local control as combined IVB/EBRT in pts with FIGO stage I and II disease. LVSI deserves more attention as a prognostic factor and these pts may require a combined local and systemic therapy.


Background
Endometrial cancer (EC) is the most common gynecologic malignancy. While therapy guidelines are widely established and the prognosis is generally favorable, optimal treatment remains controversial. In particular the conductance of systematic lymphadenectomy (LNE) and the role of external beam radiotherapy (EBRT) in early stages have been actively disputed, since both therapy modalities are known to cause substantial morbidity. Current guidelines recommend systematic pelvic and paraaortic LNE with investigation of at least 15 pelvic and 10 paraaortic lymph nodes from FIGO IB onward [1,2]. However, clinical practice differs among surgical centers and many patients are spared LNE or only undergo sampling of suspicious pelvic lymph nodes commonly omitting paraaortic lymph nodes. For radiation oncologists it can be challenging to recommend the appropriate adjuvant therapy for these patients (pts), especially when they present with additional risk factors, such as grade-3-histology or lymphovascular space invasion (LVSI). We therefore designed this retrospective analysis to elucidate the clinical outcome in the pts treated with adjuvant radiotherapy (RT) at our department and further to investigate the role of LNE and known prognostic factors.

Methods
Between 2004 and 2012 we performed adjuvant RT in 322 women with endometrioid EC at our department. All pts were included in this retrospective study, which was approved by the ethics committee of the University of Heidelberg. Due to its retrospective and blinded design consent was not required. By revision of the electronic patient charts we collected detailed information on stage, grading, resection status, LVSI, primary surgical therapy, adjuvant RT and additional adjuvant chemotherapy (ChT). FIGO 2009 classification was used for staging and patients were reclassified if necessary. Survival analysis was done for local recurrence free survival (LRFS), distant metastases free survival (DMFS) and overall survival (OS). LRFS was considered to be the time between first diagnosis and first recurrence within the irradiation field. DMFS was calculated as the time from first diagnosis until distant relapse. OS was calculated from date of first diagnosis until death from any cause. Survival was plotted according to Kaplan and Meier. The Log-rank test was used for univariate analysis and Cox proportional hazard model was used for multivariate analysis. The Chi-square test was used to illustrate heterogeneity among treatment groups. A p-value ≤ .05 was considered statistically significant. Statistical analysis was performed with SPSS 22.0 for Windows.
Primary surgical therapy consisted of hysterectomy and bilateral salpingo-oophorectomy in all patients. Pelvic and paraaortic LNE was conducted in 98 pts Only patients with stage III and IV disease received ChT. A detailed overview on surgical lymph node staging and radiotherapy for FIGO stages I-IV as well as for FIGO stage I according to risk stratification [3] is given in Table 2. Patients with intermediate risk were statistically significantly more often treated with combined EBRT/IVB when surgical lymph node staging was omitted (p = .009, Table 2).
Five-year local control in stage I was 96.0 % after sole IVB and 90.8 % after combined IVB/EBRT (p = .105).

Discussion
The conductance of systematic LNE, especially in early stage EC, has been actively disputed over the past years [4]. Two randomized trials did not find a survival benefit from systematic pelvic LNE in EC [5,6]. Arguable, in Table 3 Sites of recurrences both studies paraaortic LNE was omitted, the no-LNE groups showed a high proportion of low-risk pts and adjuvant therapy was not standardized. The Japanese SEPAL trial showed a significantly prolonged OS in intermediate-and high risk patients who underwent combined pelvic and paraaortic LNE, whereas OS was not prolonged in low-risk pts [7]. The uncertainty over the therapeutic value of systematic LNE has led to a point where surgical practice differs among centers substantially and many surgeons prefer an individual risk assessment over the systematic LNE recommended in current guidelines [1,2]. A recent report from Bogani et al. describes the clinical practice regarding LNE at the Mayo Clinic in Minnesota [8]. The colleagues from Rochester only perform systematic pelvic and paraaortic LNE in pts with >50 % myometrial invasion, nonendometrioid histology or both. Additionally paraaortic LNE is conducted when positive pelvic lymph nodes are found, while pelvic LNE is conducted based on involvement of the uterine cervix, undifferentiated grading and tumor diameter > 2 cm. This practice seems to be supported by a recent Surveillance, Epidemiology, and End Results database analysis from Vargas et al. on risk factors for lymph node metastases in EC [9]. The cohort of women treated with adjuvant RT at our department was also very heterogeneous regarding surgical staging and only a minority underwent comprehensive surgical staging as demanded in current guidelines; however we were unable to detect any statistically significant benefit from surgical lymph node investigation for all 3 endpoints (LRFS, DMFS and OS) in our current study. This was surprising, since we were able to demonstrate a significantly prolonged OS after LNE in pts with type II EC previously [10]. The groups of pts with stage IIIA and IIIB as well as stage IV disease in our current study were too small for LNE-and RT-stratified subgroup analysis ( Table 2), but for patients with stage I and II disease we were able to confirm these results in separate analyses. Women with intermediate risk more often received combined IVB/EBRT when LNE was omitted in our study; this, however, did not have a statistically significant influence on survival as well. Our results are generally supported by the findings of the PORTEC trials where no routine LNE was required [11,12]. We believe that surgical lymph node staging may be omitted in pts with type I EC in stages I-II and we further believe that women who did not receive systematic LNE and are without clinical suspicion of regional lymph node metastases may be treated as if they had undergone comprehensive surgical staging. We found our results regarding local control, diseasefree and overall survival to generally be in line with a recent retrospective study on postoperative RT from Switzerland [13]. Several randomized trials have investigated the role of adjuvant RT in EC [11,[14][15][16]. All showed a significantly improved local control; even in the treatment of recurrences [17]. Since the Norwegian trial from Aalders et al. [14], published in 1980, efforts have been made to reduce RT-associated morbidity by defining subgroups of pts which do not require EBRT and may benefit from IVB alone [12,[18][19][20][21][22]. The PORTEC-2 trial proved that IVB alone provides similar local control rates as combined IVB/EBRT treatment in pts with higher risk profile in early stages while toxicity is substantially reduced [12,23]. The women with FIGO stage I and stage II disease in our study cohort also had an equally good local control and OS after sole IVB and combined IVB/EBRT. Far too few patients received additional adjuvant ChT to be able to draw any conclusions on its effect on survival.
In multivariate analysis age over 66 years was associated with a shorter OS (p = .017). This is in agreement with a study from Benedetti Panici et al. who reported a reduced overall and cancer specific survival in women over 65 years [24]. Other independent prognostic factors for OS in our analysis were pT stage (p = .029), lymph node status (p = .003) and grading (p = .011). For local control we found resection status (p = .01) to be an independent prognostic factor. Interestingly LVSI was the only independent prognostic factor that was found in all 3 endpoints (LRFS [p = .014], DMFS [p = .049], OS [p = .008]) and it was the only independent prognostic factor for DMFS. Previous studies have already reported on the prognostic relevance of LVSI in EC or considered the presence of LVSI a feature of higher risk [13][14][15][25][26][27]. We believe that LVSI deserves more attention as a prognostic factor and that these women may be in need for a combined local and systemic treatment approach.
The observed acute GI toxicities were moderate and in line with a previous report, considering that intensity modulated radiotherapy was available only for patients treated in more recent years [28]. The relatively large proportion (>50 %) of documented acute GU toxicities in both groups is owed to the inclusion of asymptomatic grade 1 vaginal erythema.

Conclusion
Comprehensive surgical staging is rarely performed and may be omitted in women with type I EC in stages I-II. Sole IVB provides equally good local control as combined IVB/EBRT in stages I and II. Women with LVSI may be in need for a combined local and systemic therapy and LVSI should be included as a major risk factor in future randomized trials.