Impact of adjuvant radiotherapy on the survival of women with optimally resected stage III endometrial cancer in the era of modern radiotherapy: a retrospective study

Patients with surgically staged endometrial cancer treated between 2010 and 2018 within the multi-institution National Taiwan University Hospital Healthcare System (including the National Taiwan University Hospital, the affiliated Yun-Lin branch, and Hsin-Chu branch) were retrospectively investigated. Staging was performed according to the 2009 International Federation of Gynecology and Obstetrics (FIGO) staging system for endometrial cancer [18]. A total of 1273 patients with surgically staged endometrial cancer were identified, among whom 204 had FIGO stage III. Women who were administered chemotherapy or radiotherapy before surgery, received surgery with a palliative intent, or had sarcoma or carcinosarcoma were excluded; hence, 161 patients with stage III endometrial cancer who underwent optimal resection were ultimately included.

Staging surgery included total abdominal hysterectomy, bilateral salpingo-oophorectomy, peritoneal washings, and either selective or systematic pelvic lymphadenectomy. The para-aortic lymph nodes were sampled or dissected from the area between the aortic bifurcation and inferior mesenteric artery in patients with elevated serum cancer antigen 125, myometrial invasion > 50%, extrauterine spread, or para-aortic lymph nodes > 1 cm in diameter as identified on preoperative magnetic resonance imaging (per the Korean Gynecologic Oncology Group 2014 criteria). Bulky nodes were removed by dissection whenever possible. An omental biopsy or omentectomy was performed at the discretion of the individual surgeon based on the extent of disease. The patients had no residual disease by the end of primary surgery; the median number of dissected lymph nodes was 12 (range: 0–40).

Chemotherapy regimens were administered according to the physicians’ preferences; the most common treatments were platinum-based regimens with the majority (81%) receiving paclitaxel and platinum (carboplatin or cisplatin). Other regimens included cisplatin plus doxorubicin (10%), and cisplatin plus epirubicin (9%). The majority of patients received 6 cycles of chemotherapy. Women who received both adjuvant chemotherapy and radiotherapy completed all chemotherapy courses before radiotherapy (34%) or in conjunction with radiotherapy in a sandwich pattern (66%).

For optimally resected stage III endometrial cancer, the adjuvant radiotherapy was directed at sites of known or suspected tumor involvement, and included EBRT and/or brachytherapy [19]. All women who underwent adjuvant radiotherapy were treated with EBRT, and vaginal brachytherapy was also used for most (84%) women. Our health-care system adopted modern radiotherapy including IMRT since 2010 and VMAT since 2015, which helped minimize the dose to the normal organs on the basis of adjuvant radiotherapy [20, 21]. The dose of EBRT was 5040 cGy over 6 weeks, 5 days per week, with daily fractions of 1.8 Gy. Pelvic radiotherapy targeted the common, external, and internal iliac lymph node regions, upper 3 cm of the vagina, and the paravaginal soft tissue lateral to the vagina in accordance with the updated delineation consensus for gynecologic malignancy [11]. The presacral lymph nodes were irradiated in patients with cervical involvement. In patients with multiple positive pelvic nodes or documented para-aortic lymph node disease, extended-field radiotherapy that encompassed the pelvic volume and also targeted the entire common iliac chain and para-aortic lymph node region was considered. A boost dose of 5–10 Gy was also considered for documented extranodal extension or enlarged unresected lymphadenopathy. An example of a modern radiotherapy technique with its associated isodose curves is shown in Fig. 1. EBRT was performed using 10 MV radiation beams from the Elekta Synergy accelerator (Elekta, Stockholm, Sweden) or the Varian TrueBeam™ Radiotherapy System (Varian, Palo Alto, CA, USA) in multiple coplanar ports. The treatment position was verified weekly using cone-beam CT X-ray volume imaging [14]. Following pelvic radiation, high dose-rate brachytherapy via a vaginal cylinder was used to boost the upper two-thirds of the vagina. Brachytherapy doses of 6 Gy per fraction for 2 to 3 fractions prescribed to the vaginal mucosa were used, and they were delivered using the “Nucletron HDR” Ir-192 remote afterloading technique [22, 23].

Statistical analyses were performed using the Statistical Package for Social Sciences for Windows, version 22.0 (IBM, Armonk, NY, USA). Differences in the distributions of variables among the treatment groups were analyzed using the chi-squared test. Acute and late toxicities were rated according to the Common Terminology Criteria for Adverse Events version 4.0. Survival data were confirmed with the Cancer Registry Medical Information Management Office in the health-care system. All patients were followed for at least 3 months for the first 2 years and every 4–6 months thereafter until recurrence or death [2, 23]. Locoregional recurrence was defined as failure in the pelvic regions or evidence of para-aortic lymphadenopathy below the T12–L1 interspace. Distant metastasis was defined as disease relapse outside the locoregional area as detected pathologically, cytologically, or radiologically. All events were calculated from the date of treatment completion. Analysis was conducted using the follow-up data accumulated as of September 30, 2019. Actuarial estimates of recurrence-free survival (RFS) and OS were calculated using the Kaplan-Meier method and compared using the log-rank test. All prognostic variables found to be significant on univariate analysis were subjected to multivariate analysis using the Cox proportional hazards regression model. A p-value ≤0.05 was considered statistically significant.

A total 161 optimally resected patients were investigated (52 with stage IIIA, 9 with stage IIIB, and 100 with stage IIIC disease); none had residual disease upon completing primary surgery. The patient and tumor characteristics are shown in Table 1. The median age at diagnosis was 57 years, and the predominant histologic subtype was endometrioid (85%). Lymphovascular space invasion was present in nearly three-fourths (74%) of the tumors, and nearly one-third (32%) had FIGO grade 3 histology. Positive peritoneal washing cytology was observed in nearly one-fourth (24%) of the patients.

One hundred fifty-four patients (96%) received an adjuvant treatment; 67 (42%) received chemotherapy alone, 29 (18%) received adjuvant radiotherapy alone, and 58 (36%) underwent chemoradiotherapy. In the 87 patients who received adjuvant radiotherapy with or without chemotherapy, IMRT was used in 45 (52%) patients and VMAT was used in 42 (48%) patients. Of the remaining 7 patients, 4 decided against adjuvant therapy and 3 had comorbidities that precluded such treatments. The relationships between the patients’ characteristics and type of postoperative adjuvant treatment received are described in Table 2. There was no significant difference in the type of administered treatment among patients grouped according to FIGO stage, tumor histology, tumor grade, lymphovascular space invasion, positive peritoneal cytology, or number of positive lymph nodes. However, elderly patients were less frequently treated with combined chemoradiotherapy than were their younger counterparts (p = 0.046).

Regarding adjuvant radiotherapy-related gastrointestinal side effects, 23% of patients had acute grade 2 diarrhea, and 2% had acute grade 3 diarrhea; 5% had grade 2 late toxicity with moderate diarrhea requiring medications, and 2% developed grade 3 radiation proctitis that were successfully treated by colonoscopy argon plasma coagulation. None had acute or late grade 4+ gastrointestinal symptoms. After a median follow-up of 44 months (range: 3–103 months), 140 patients (87%) were alive, whereas 21 patients (13%) had died; the majority of deaths (95%) were attributed to cancer progression. Fifty-one patients had tumor recurrence, which was locoregional in 26 (vaginal stump, pelvic, or para-aortic recurrence) and distant in 41 (lung, liver, bone, brain, distant lymphadenopathy, or peritoneal carcinomatosis); moreover, 16 patients had both locoregional recurrence and distant metastasis. Administration of any type of adjuvant treatment was associated with a longer 5-year RFS (67% vs. 0%, p = 0.014, Supplementary Figure 1a) and OS (84% vs. 66%, p = 0.044, Supplementary Figure 1b). The type of adjuvant treatment did not influence the risks of tumor recurrence in patients with optimally resected stage III disease; the 5-year RFS was 61% in those who underwent adjuvant chemotherapy alone, 73% in those who received adjuvant radiotherapy alone, and 79% in those who received combined chemoradiotherapy (p = 0.172, Fig. 2a). Likewise, the 5-year OS was 80% for patients treated with adjuvant chemotherapy alone, 85% for those treated with adjuvant radiotherapy alone, and 86% for those who received combined chemoradiotherapy (p = 0.390, Fig. 2b).

On the univariate analysis, tumor stage (IIIA or IIIB vs. IIIC), presence of lymphovascular space invasion, positive peritoneal washing cytology, or ≥ 2 positive lymph nodes were not significantly associated with an increased risk of tumor recurrence or death, whereas older age was associated with an increased risk of tumor recurrence (hazard ratio [HR] 2.14, p = 0.008) but not of death (Table 3). Non-endometrioid histology and grade 3 tumor status were associated with increased risks of tumor recurrence (HR: non-endometrioid histology: 5.42; grade 3 tumor status: 1.74, both p < 0.001) and death (HR: non-endometrioid histology: 4.34, p = 0.001; grade 3 tumor status: 1.49, p = 0.043).

Moreover, the type of adjuvant treatment was associated with clinical outcomes. Patients who received adjuvant radiotherapy alone or combined with chemotherapy experienced a longer 5-year RFS than those who did not receive radiotherapy (75% vs. 61%, p = 0.046, Fig. 2c) and also showed a trend toward a longer 5-year OS rate (85% vs. 80%, p = 0.065, Fig. 2d). Meanwhile, patients who received adjuvant chemotherapy alone or combined with radiotherapy had similar survival periods as those who did not receive chemotherapy (5-year RFS: 69% vs. 73%, p = 0.838, Fig. 2e; 5-year OS: 82% vs. 85%, p = 0.965, Fig. 2f).

On the multivariate analysis, non-endometrioid histology increased the risk of tumor recurrence (HR 2.95, 95% confidence interval [CI] 1.32–6.61, p = 0.009), whereas adjuvant radiotherapy alone or combined with chemotherapy decreased the risk of tumor recurrence (HR 0.62, 95% CI 0.31–0.98, p = 0.042) (Table 4).

A subgroup analysis was performed to identify patients who might most benefit from adjuvant radiotherapy (Fig. 3). Adjuvant radiotherapy alone or combined with chemotherapy decreased the risk of tumor recurrence among patients aged > 60 years (p = 0.038) as well as those with endometrioid histology (p = 0.045), lymphovascular space invasion (p = 0.031), and ≥ 2 positive lymph nodes (p = 0.044). However, adjuvant radiotherapy did not significantly reduce the risk of recurrence among patients with grade 3 tumors or those with positive peritoneal washing cytology.