Abstract
Background/Aim: To assess the patterns of recurrence of node-positive endometrial cancer patients. Patients and Methods: This investigation assessed 82 patients who received different postoperative treatments. Results: Recurrence developed in 36 patients after a median time of 13.5 months, and involved the vagina, pelvic nodes, para-aortic nodes and distant sites in 5, 8, 16 and 17 patients, respectively. Five-year progression-free survival (PFS) and 5-year overall survival (OS) were 51.1% and 59.8%. PFS and OS were significantly better for endometrioid than for non-endometrioid tumors. There was a trend towards a better outcome for patients who underwent chemotherapy±radiotherapy compared to those who received radiotherapy alone. Among the former, there was a better 5-year PFS (65.8% versus 33.7%, p=0.038) in patients who received platinum/paclitaxel-based regimens compared to those who received platinum-based chemotherapy. Conclusion: Disease recurred in 43.9% of patients, and platinum/paclitaxel-based chemotherapy plus radiotherapy appeared to be the best adjuvant treatment.
The standard surgery of endometrial cancer consists of total hysterectomy with bilateral salpingo-oophorectomy carried out with open or mini-invasive approach, whereas the role and the extension of retroperitoneal node dissection is still debated (1-5). Node metastases can be found in approximately 10% of women with apparently early stage endometrial cancer at the preoperative work-up, and systematic lymphadenectomy should be an integral part of the International Federation of Gynecology and Obstetrics [FIGO] staging system, useful for the definition of prognosis and planning of adjuvant therapy (6-8). However, the therapeutic benefit of this surgical procedure is still uncertain. Two randomized trials failed to detect any advantage in progression-free survival (PFS) and overall survival (OS) for patients who underwent pelvic lymphadenectomy compared to those who did not, although these trials had been criticized for several biases (9-11). Moreover, the women in the lymphadenectomy arm experienced a higher risk of surgery-related systemic morbidity and lymphoedema/lymphocyst formation. As suggested by the Mayo Clinic historical surgical algorithm (12, 13), lymphadenectomy should be carried out in patients with apparent stage I endometrioid endometrial cancer and grade 3 on preoperative biopsy and/or myometrial invasion ≥50% on intraoperative frozen sections of the uterine specimen as well as in patients with non-endometrioid tumors, whereas this surgical procedure can be omitted in endometrioid endometrial cancer patients with no myometrial invasion (regardless of grade or tumor diameter) or with grade 1-2, <50% myometrial invasion and tumor diameter ≤2 cm. The sentinel lymph node biopsy is considered to be experimental by the ESMO-ESGO-ESTRO Consensus Conference on endometrial cancer (3), whereas it has been included in the National Comprehensive Cancer Network (NCCN) guidelines for this malignancy (14).
Among the 7,990 surgically staged endometrial cancer patients reported in the FIGO annual Report n. 26, 356 (4.5%) had stage IIIc disease, and their 5-year OS was 57.3% (6). Radiotherapy and chemotherapy have been widely used as postoperative treatments, but no definitive conclusion can be drawn on the optimal adjuvant therapy in this clinical setting (3, 11, 13, 15-34).
Treatment modalities.
The purpose of this retrospective investigation was to analyze the rates, sites and times of recurrences and the clinical outcome of patients with FIGO stage IIIc1-IIIc2 endometrial cancer treated at two Italian Gynecological Oncology centers.
Materials and Methods
This retrospective investigation assessed 82 patients who underwent peritoneal washing, extrafascial (Piver-Rutledge class I) or modified radical (Piver-Rutledge class II) hysterectomy, bilateral salpingo-oophorectomy and pelvic plus aortic lymphadenectomy for endometrial cancer and who had histologically proven positive nodes at the Department of Gynecology and Obstetrics of the University of Pisa between 1993 and 2016 and the Department of Gynecology and Obstetrics of the University of Turin (Mauriziano Hospital) between 2003 and 2016. The patients who did not undergo aortic lymphadenectomy had negative computed tomography (CT) findings for node aortic involvement. Abdomen-pelvis CT scan was performed two to three weeks before surgery, and aortic nodes were defined negative when their short axes were <5 mm. Some of the patients treated in Pisa had been included in a previous paper of our group (35).
Primary sites of failure.
Sixty-seven women were operated via laparotomy and 15 via minimally invasive approach (laparoscopy or robotics). Women with carcinosarcoma were excluded from the present analysis. According to Mariani et al. (13), an adequate pelvic lymphadenectomy was defined as the removal of >10 pelvic nodes, and an adequate para-aortic lymphadenectomy was defined as the removal of ≥5 para-aortic nodes.
Patients were staged retrospectively according to the FIGO 2009 classification (7).
The architectural grade was defined as follows: G1, ≤5% of non-squamous or non-morular solid growth pattern; G2, 6-50% of non-squamous or non-morular solid growth pattern; and G3, >50% of non-squamous or non-morular solid growth pattern. Notable nuclear atypia, inappropriate for the architectural grade, raised the grade of G1 or G2 tumor by one.
Lymph-vascular space involvement (LVSI) was defined as the presence of tumor cells within or attached to the wall of a blood vessel or lymphatic space using morphological and immunohistochemical analyses.
Post-operative treatment was established on the basis of pathological findings on surgical specimens, patient age and general conditions. However, adjuvant therapy has been changed in the two centers over the long interval time of the study.
External-beam radiotherapy (EBRT) was performed with a 15-18 MV beam, and a 45-50.4 Gy dose was given to the pelvis in daily fractions of 1.8 Gy in 5-6 weeks. The target volume was outlined on a CT scan. When performed, 45 Gy para-aortic irradiation was planned in daily fractions of 1.8 Gy in 5 weeks. Vaginal cuff high-dose rate brachytherapy (BCT) was added in selected cases with isthmus or stromal cervical involvement after EBRT. The prescribed dose was 10-15 Gy in 5 Gy fractions. Rectal and bladder doses were estimated from dose volume histograms on CT-based plans and were evaluated to the dose points specified by the International Commission on Radiation Units and Measurements.
Adjuvant chemotherapy consisted of platinum-based regimens. EBRT was delivered sequentially after chemotherapy in patients who received both adjuvant treatments.
Treatment and clinical outcome of patients with recurrent disease.
Follow-up procedures are reported in a previous paper (36). All the patients were periodically followed until September 2017 or until death. The median follow-up of survivors was 66.9 months (range=8.0-207.8 months).
Statistical analysis. Age, FIGO stage (IIIC1 versus IIIC2), tumor grade (G1-2 versus G3), histology (endometrioid versus non-endometrioid), myometrial invasion (<50% versus ≥50%), LVSI, cervical involvement, adnexal involvement, type of radical hysterectomy (I versus II), adjuvant treatment (chemotherapy+ EBRT versus chemotherapy versus radiotherapy alone), and chemotherapy regimen (platinum/ paclitaxel-based versus platinum-based) were analysed for association with PFS and OS. Peritoneal, hematogenous, and node recurrences outside the retroperitoneal area (i.e. mediastinal) were considered as distant failures.
SAS statistical package (release 8.2; SAS Institute, Cary, NC, USA) was used for the computations.
The time from surgery to the detection of recurrence was defined as PFS. The time from surgery to death or last observation was defined as OS. The cumulative probability of PFS and OS were estimated by the product-limit method. The log-rank test was used to compare the homogeneity of PFS and OS functions across strata defined by categories of prognostic variables.
Results
At presentation, median age of patients was 64 years (range=36-85 years). FIGO stage was IIIC1 in 54 (65.9%) and IIIC2 in 28 (34.1%) women. Tumor grade was G1 in 5 (6.1%), G2 in 33 (40.2%), and G3 in 44 (53.7%) patients, respectively. Histological type was endometrioid in 58 (70.7%) and non-endometrioid in 24 (29.3%) women (serous, 14; clear cell, 4; undifferentiated, 6). Myometrial invasion ≥50%, LVSI, cervical involvement, and adnexal involvement were found in 70 (85.4%), 58 (70.7%), 24 (29.2%), and 18 (21.95%) patients, respectively. Treatment modalities are reported in Table I.
Pelvic lymphadenectomy was performed in all 82 women, and positive pelvic nodes were found in 77 (93.9%). The median number of metastatic pelvic nodes was 2 (range=1-21 nodes). Para-aortic lymphadenectomy was performed in 32 of these 77 patients, and positive para-aortic nodes were found in 21 (65.6%).
Para-aortic lymphadenectomy was performed in 38 out of the 82 patients, and positive para-aortic nodes were found in 26 (68.4%). The median number of metastatic para-aortic nodes was 2 (range=1-17 para-aortic nodes). The pelvic nodes were positive in 21 of these 26 patients (80.8%).
Postoperative treatment consisted of chemotherapy in 21 women (25.6%), chemotherapy followed by EBRT in 50 (61.0%), EBRT in 10 (12.2%), and BCT alone in one (1.2%).
Tumor relapsed in 36 out of the 82 patients (43.9%), with a median time to recurrence of 13.5 months (range=4.1 to 35.2 months). Overall, recurrent disease involved vagina in 5 (6.1%), pelvic nodes in 8 (9.8%), para-aortic nodes in 16 (19.5%), and distant sites in 17 (20.7%), respectively, of the 82 patients (Table II).
Recurrent disease had a very poor prognosis (Table III). Eleven of the 12 (91.7%) patients with distant failure, 12 of the 14 patients (85.7%) with retroperitoneal node failure, all the 6 patients (100%) with multiple site failure, and all the 4 patients (100%) with isolated vaginal failure died of disease. Three of these latter had previously undergone adjuvant pelvic EBRT (followed by brachytherapy in one case).
Clinical outcome of patients according to the prognostic variables.
Five-year PFS and 5-year OS of the 82 patients were 51.1% and 59.8%, respectively. PFS and OS were significantly better in patients with endometrioid than in those with non-endometrioid tumors (p<0.0001 and p=0.001) (Table IV). There was a trend towards a better PFS and OS in patients who underwent adjuvant chemotherapy or adjuvant chemotherapy followed by EBRT compared to those who received radiotherapy alone.
Among the patients who received adjuvant chemotherapy with or without EBRT, there was a significantly better PFS (p=0.0038) and a trend to a better OS for the patients who received platinum/paclitaxel-based regimens compared to those who received platinum-based chemotherapy.
Discussion
Although node involvement is a strong prognostic variable for endometrial cancer, the therapeutic potential of lymphadenectomy in this malignancy is still a matter of debate (8-10, 15, 28, 30, 31, 33, 37, 38). According to the recommendations of ESMO-ESGO-ESTRO Consensus Conference regarding apparent stage I endometrioid endometrial cancer, lymphadenectomy: i) is not indicated for low-risk disease (G1-2 and myometrial invasion <50%); ii) can be considered for staging purposes for intermediate-risk disease (myometrial invasion >50% or G3; iii) and should be indicated for high-risk disease (G3 with myometrial invasion >50%) (3). Moreover, lymphadenectomy is recommended for clinical or intra-operative stage II endometrioid endometrial cancer as well as for apparent stage I non-endometrioid endometrial cancer. When performed, systematic removal of pelvic and para-aortic nodes should be carried out up to renal veins. Two retrospective studies comparing two nodal assessment approaches in patients with endometrioid endometrial cancer treated at the Majo Clinic and at the Memorial Sloan Kettering Cancer Center showed that the sentinel lymph node algorithm and lymphadenectomy algorithm had the same oncologic outcome in low-risk patients and similar nodal metastasis detection rates in high-risk patients (39, 40). Other studies have confirmed that sentinel node mapping can represent a safe, alternative option to minimize lymph node dissection without compromising surgical staging of endometrial cancer (41, 42). However, according to the ESMO-ESGO-ESTRO guidelines, sentinel node biopsy is still experimental in endometrial cancer (3).
Studies investigating adjuvant chemotherapy in patients with node-positive endometrial cancer.
Data from literature have shown that 5-year-OS ranges from 53.9% to 81.0% in stage IIIC, and from 61 % to 86.4% in Stage IIIC1 and from 42.3% to 66.3% in stage IIIC2 (6, 16, 17, 25, 27-29, 33).
Many authors have investigated the role of adjuvant radiotherapy and/or chemotherapy in early-stage, high-risk or advanced-stage endometrial cancer (11, 18-22, 33, 34, 43). Only few papers have analyzed the results of postoperative radiotherapy selectively in stage IIIC disease (13, 27, 32). The analysis of the 2177 women with stage IIIC endometrial carcinoma (IIIC1=1363, IIIC2=658) included in the Surveillance, Epidemiology, and End Results (SEER) database from 2004 to 2012 found that 1248 (60%) had undergone adjuvant radiotherapy (32). The 3-year OS was 80.5% and 67.6%, for the patients who received radiotherapy and for those who did not, respectively (p<0.001), and the use of this adjuvant treatment was an independent prognostic variable for OS [Hazard ratio (HR)=0.61, 95% Confidence interval (CI)=0.51-0.74] at multivariate analysis.
Recently adjuvant chemotherapy has been more and more employed in endometrial cancer. Meta-analysis of four randomized trials comparing chemotherapy versus radiotherapy after surgery in 1326 patients with high-risk early or advanced disease showed that chemotherapy was associated with a relative risk (RR) of death at 5 years of 0.87 (95%CI=0.76-0.99) (19-21, 44, 45). However, the statistical significance was lost, after omission of the GOG 122 study, which was not a pure adjuvant trial because it included also patients with residual disease up to 2 cm.
Four studies assessing adjuvant chemotherapy alone in patients with node positive endometrial cancer reported 5-year OS rates ranging from 48.1% to 84.8% (Table V).
Young et al. (30) reported that tumor relapsed in 8 of the 25 (32.0%) with stage IIIC1 disease and in 7 of the 18 (38.9%) with stage IIIC2 disease. Failure involved pelvic area, para-aortic area, pelvic plus para-aortic area, and distant sites in 1 (4%), 3 (12%), 3 (12%), and 2 (8%) of stage IIIC1 patients, and, respectively, in 1 (5.6%), 2 (11.1%), 1 (5.6%), and 3 (16.7%) of stage IIIC2 patients.
Studies investigating adjuvant chemotherapy, radiotherapy and chemotherapy+radiotherapy in patients with node-positive endometrial cancer.
Table VI shows the studies investigating adjuvant chemotherapy, radiotherapy and chemotherapy plus radiotherapy in patients with node-positive endometrial cancer.
In the study of Klopp et al. (23) tumor failure involving vagina, pelvic area, para-aortic area, and distant sites developed in 4 (22.2%), 4 (22.2%), 2 (11.1%) and 2 (11.1%) of the 18 patients treated with systemic therapy without external radiotherapy, versus 5 (10.0%), 1 (2.0%), 3 (6.0%), and 13 (26.0%) of the 50 patients who received external radiotherapy with or without systemic therapy.
Adjuvant platinum-based chemotherapy interposed with radiotherapy was administered to 43 patients with advanced endometrial cancer (24). The majority had stage IIIC disease (63%). Twenty-one patients (49%) relapsed at a median time of 17 months, and recurrent disease was local in 2 patients (9.5%), distant in 18 (85.7%), and local+distant in 1 patient (4.8%). Sixty-two (94%) out of 66 patients assessed by Lee and Viswanathan (25) received adjuvant radiotherapy with or without chemotherapy. Seventeen patients (25.8%) recurred after a median time of 21 months. PFS and OS were better for patients treated with chemotherapy plus radiotherapy compared to those who received radiotherapy alone.
The sites of failures were vaginal in 10 (14%), pelvic in 7 (9%), extra-pelvic in 52 (70%), and not specified in 5 (7%) of the 262 patients with stage IIIC disease treated by Secort et al. (26), with no significant differences in overall recurrence rates according to adjuvant therapy. Fleming et al. (31) found a significantly lower recurrence rate in patients receiving radiotherapy with or without chemotherapy compared to those treated with chemotherapy alone.
In a large Taiwanese study, 174 of 541 patients (32.2%) with advanced endometrioid endometrial cancer developed a recurrence, which was local in 47 (27.0%), distant in 79(45.4%), and both local and distant in 48 (27.5%) (33). Paclitaxel-based multimodality treatment was an independent prognostic factor for better PFS (HR=0.608; 95%CI=0.403-0.916) and OS (HR=0.482; 95%CI=0.310-0.749).
Recent randomized phase III trials showed that addition of adjuvant chemotherapy to radiotherapy prolongs PFS compared to radiotherapy alone in patients with high-risk profile (22, 34). For instance, in the PORTEC 3 trial the combined treatment significantly improved 5-year PFS by 11% (69.3% versus 58.0%, p=0.032) (34).
In our study, endometrial cancer relapsed in 43.9% of 82 patients, after a median time of 13.5 months, and recurrent disease involved vagina in 6.1%, pelvic nodes in 9.8%, para-aortic nodes in 19.5%, and distant sites in 20.7% of the patients, respectively. The rates, times and sites of relapses are in agreement with the literature (23, 24, 26, 30, 33). In our series, PFS and OS were significantly better in patients with endometrioid tumors than in those with non-endometrioid tumors. Similarly, Sueoka et al. (28) detected a 5-year OS of 90.2% for endometrioid carcinomas versus 56.7% for non-endometrioid carcinomas (p=0.0016), and Young et al. (30) reported a better 3-year PFS and 3-year OS for endometrioid than for non-endometrioid tumors (92.4% versus 58.0%, p=0.001, and, respectively, 97.2% versus 65.8%, p=0.002). Conversely other authors failed to detect a prognostic relevance for histologic type (16, 17). In the present investigation, depth of myometrial invasion, tumor grade, LVSI, cervical involvement and adnexal involvement were not related to the PFS and OS. However, literature data about the prognostic relevance of these variables in stage IIIC endometrial cancer are controversial. Several authors reported that myometrial invasion and adnexal involvement did not correlate with the clinical outcome (16, 17, 25). High tumor grade was associated with worse prognosis in the study of Fujimoto et al. (17) but not in others (16, 25), cervical involvement got worse PFS and OS in the study of Lee et al. (25) but not in others (16, 17), and LVSI correlated with unfavorable outcome in the series of Watari et al. (16) but not in others (17, 25).
In conclusion, patients with FIGO stage IIIC1-IIIC2 endometrial cancer relapsed in 43.9% of the cases, and distant sites and para-aortic nodes represented the common sites of recurrence. There was a trend towards a better outcome in patients who underwent adjuvant chemotherapy or adjuvant chemotherapy followed by EBRT compared to those who received radiotherapy alone. Among the patients who received adjuvant chemotherapy with or without EBRT, there was a significantly better 5-year PFS (65.8% versus 33.7%, p=0.038) and a trend to a better 5-year OS (66.7% versus 52.6%) for the patients who received platinum/paclitaxel-based regimens compared to those who received platinum-based chemotherapy. However, the retrospective, non-randomized nature of the study and the limited number of patients did not allow to draw any conclusions regarding the impact of adjuvant treatment on the pattern of recurrences.
- Received December 28, 2017.
- Revision received January 23, 2018.
- Accepted January 24, 2018.
- Copyright© 2018, International Institute of Anticancer Research (Dr. George J. Delinasios), All rights reserved
