Pattern of practice for postoperative management of endometrial cancer in Korea: a survey by the Korean Gynecologic Oncology Group and the Korean Radiation Oncology Group (KGOG 2028-KROG 2104)

Endometrial cancer (EC) is the seventh most common malignancy in females in Korea, and its incidence rapidly increased from 5.6 per 100,000 in 2010 to 8.8 per 100,000 in 2019 [1]. EC is treated primarily by surgical resection and various adjuvant strategies, such as pelvic radiotherapy (RT) and systemic chemotherapy (CTx). The risk assessment for recurrence of early-stage EC generally considers various clinicopathological factors, such as the histological grade, myometrial invasion, lymphovascular invasion, tumor location, and patient age. In addition to the traditional risk classification, including low, intermediate, and high-risk groups, intermediate risk has been subdivided into high–intermediate risk and low–intermediate risk according to the Gynecologic Oncology Group (GOG)-99 trial [2]. Advanced-stage EC also comprises a heterogeneous group of diseases ranging from direct local extension of the disease to nodal or distant metastasis.

To support evidence-based clinical guidance on adjuvant treatment for EC, a series of large-scale clinical trials has been performed by the GOG and the Postoperative Radiation Therapy for Endometrial Carcinoma (PORTEC) study group. The results of 3 landmark studies have been recently published by the GOG-249 [3], PORTEC-III [4], and GOG-258 [5] trial groups (Table 1). These studies may affect the clinical decisions of gynecologic cancer experts, and there is a need to investigate how this has influenced current patterns of practice. The Korean Gynecologic Oncology Group (KGOG) and the Korean Radiation Oncology Group (KROG) are organizations of gynecologic oncologists and radiation oncologists (ROs) that were initiated in 2002 and 2001, respectively, for clinical multi-institutional collaborative study and research. This study was designed with this infrastructure to identify the current patterns of practice of adjuvant treatment for EC in Korea by distributing a survey to the KGOG and KROG members. Importantly, we compared the patterns of practice between gynecologic cancer surgeons (GYNs) and ROs with an intent to identify the discrepancies between the experts and improve patient care for EC in Korea.

Table 1
Summary of recent clinical trials according to the Participant, Intervention, Comparison, and Outcome assessment form

• Click for larger image
• Click for full table
• Download as Excel file

A mail survey involving an electronic questionnaire consisting of 25 questions regarding the status of adjuvant treatment and case scenarios was distributed to members of the KGOG and KROG. The questionnaire was designed by the principal investigator (K.J.Y.) and was approved by the Disease Committee of Uterine Endometrial Cancer of the KGOG (KGOG-2028) and the gynecologic cancer study branch of the KROG (KROG-2104). In accordance with Declaration of Helsinki, the institutional review boards of National Cancer Center granted an exemption for this study. As a general questionnaire for clinical decision, intended to investigate the up-to-date clinical practice after the recent GOG and PORTEC studies, the 13 questions were composed of 6 demographic and 7 clinical trial questions (Table S1). Because the risk assessment of recurrence of EC includes various clinicopathologic factors, the survey with specific clinical case questions was also asked, consisted of 4 questions regarding stage I/II endometrioid EC, one regarding stage III/IV endometrioid EC, 3 regarding an unfavorable histology, and 2 regarding uterine sarcoma and carcinosarcoma (Table S2). The respondents were reminded of the results of the relevant clinical trial. The adjuvant treatment options given to the respondents included CTx alone, external beam RT (EBRT) alone, sequential use of CTx and EBRT (CTx + EBRT), vaginal brachytherapy (VBT) alone, CTx with VBT, or concurrent chemoradiotherapy (CCRT). As a last, 17 questions specific to GYNs and 16 questions specific to ROs were administered, for more detailed description of their clinical practice.

Initially, we identified 145 members in KGOG and 74 ROs who sub-specialized in gynecology oncology in KROG and mailed the questionnaire to all KGOG and KROG members in May 2021, and the last date for receipt of responses was July 13, 2021. Finally, a total of 38 GYNs and 31 ROs in 43 institutions was participated in the survey. The response rate was 26.2% (38/145) and 41.8% (31/74), respectively. One of the considerations for response rate is that some responders noted he or she responded the survey as a representative of gynecology oncology surgeons in their institution. A standardized computer software package (SPSS version 27.0; IBM Inc., Armonk, NY, USA) was used for the statistical analysis. Chi-square statistics were used to compare data between the GYN and RO groups, and a p-value <0.05 was considered significant.

In this study, we observed different opinions between GYNs and ROs regarding postoperative management of EC. Overall, the interpretations and pattern of practice for early-stage EC with reference to the results of the PORTEC-II, GOG-249, and PORTEC-III trials were similar between the 2 expert groups; however, the main areas of disagreement were the interpretation of the GOG-258 results and managing stage III/IVA endometrioid EC, unfavorable histology, and sarcomas (Fig. 3). For such cases, GYNs showed a preference for administering adjuvant CTx alone, whereas ROs advocated for RT with or without combined CTx.

Fig. 3
Comparison of responses between gynecologic surgeon and radiation oncologist regarding the need of adjuvant radiotherapy for each risk groups and histologies of endometrial cancer.
GYN, gynecologic cancer surgeon; RO, radiation oncologist.

• Click for larger image
• Download as PowerPoint slide

There may be several reasons for the differences in the opinions of GYNs and ROs. The GOG-258 trial demonstrated that CCRT and CTx improved locoregional control and reduced distant metastasis, respectively. ROs may base their opinions on the improved loco-regional control with RT, although it may not improve survival. GYNs may base their opinions on the overall survival rate and recommend systemic treatment that can be directly administered in their clinics. There is a fundamental gap in the understanding of GYNs and ROs regarding the impact of loco-regional recurrence. GYNs suggest that the reduced loco-regional recurrence with RT is less important than the reduced incidence of metastasis with CTx, because distant metastasis determines survival in a large proportion of patients. In contrast, ROs suggest that loco-regional recurrences cause severe symptoms with the potential to substantially compromise the patient’s quality of life. Furthermore, loco-regional recurrences have limited treatment options and may serve as sources for distant seeding of cancer cells.

Gynecological cancer specialists must be cautious when interpreting the results of large-scale clinical trials and translating them into clinical practice. When interpreting the results of PORTEC-II, GOG-249, and PORTEC-III trials, it is important to consider that the methods used for risk stratification and evaluation of LNs were different from the methods used in previous GOG studies. For example, lympho-vascular space invasion is the most important prognostic factor regarding regional recurrence [8, 9, 10, 11]. Even when LNs were evaluated for risk factor and staging, it was based on LN examination varied from removal of suspicious nodes, nodal sampling to partial or complete nodal dissection. Accordingly, risk groups in previous studies had heterogeneous prognostic features that failed to distinguish between loco-regional and systemic relapses. Consistent with this heterogeneity, analysis of the National Cancer Database (NCDB) regarding patients with uterine cancer who were treated between 2004 and 2017 revealed wide variation in survival outcomes according to the tumor characteristics of patients who met the pathological criteria for enrollment in GOG-249, PORTEC-3, and GOG-258 trials (i.e., survival rates of 59.9%–81.7%, 40.2%–81.8%, and 17.5%–75%, respectively) [12]. These findings suggest that the 3 recently published clinical trials included patient cohorts with heterogeneous tumor characteristics. Specifically, among patients who met the criteria for enrollment in the GOG-258 trial, the 5-year overall survival of the 7,012 patients who received CTx alone was significantly lower than the 5-year overall survival of the 8,926 women who received chemoradiotherapy (57.8% and 72.7%, respectively). The NCDB real-world data suggest that the large-scale trials were underpowered and did not represent patients with important tumor characteristics; therefore, the results of the GOG-258 trial should be interpreted with caution.

Considering this background, there is a need to understand how the results of previous clinical trials have been applied to clinical practice. After the PORTEC-1 and GOG-99 trials were published in the early 2000s, Ko et al. [13] assessed changes in the pattern of practice regarding adjuvant RT for EC in the U.S. using the Surveillance, Epidemiology, and End Results database. They compared the use of RT pre- vs. post-publication of 2 sentinel studies. The authors concluded that the utility of adjuvant RT for early-stage EC had not changed despite the proven risk reduction by adjuvant RT that was revealed in the PORTEC-1 and GOG-99 trials [13]. Although the overall use of adjuvant RT has not changed, the use of VBT has gradually increased. Naumann et al. [14] reported a markedly decreased use of pelvic RT and increased use of VBT for early-stage EC between 1999 and 2005. This trend continued until recently, when Modh et al. [15] reported a significant overall increase in the use of VBT of 17.1% during 1995–2000 to 57.1% during 2007–2012 for early-stage EC, along with a proportional decrease in the use of EBRT from 54.0% to 25.5%. In 2019, the GOG-249 trial revealed opposite results; VBT did not demonstrate superior efficacy over EBRT in high–intermediate- to high-risk EC, even with greater pelvic and para-aortic recurrence rates. Therefore, patient selection for VBT in early-stage EC would be an important issue.

In the GOG-122 randomized trial, systemic CTx was advantageous over whole-abdomen RT for locally advanced disease, and thus it became part of standard adjuvant management [7]. As revealed in the present study, the majority of GYNs have been hesitant to proceed with adjuvant RT since the GOG-122 trial. However, many GYNs do not understand that a 2,000 cGy dose of whole abdominopelvic RT represents a suboptimal eradiating dose, as well as a low therapeutic ratio, resulting in high toxicity and low tumor control [16]. IMRT decreases radiation exposure to normal organs and significantly reduces the incidence of treatment-related toxicity compared with the conventional technique [17]. As shown in this study, IMRT has become a mainstay RT technique for EC patients in Korea. In addition, if pelvic RT is omitted, CTx has a limited effect in preventing locoregional recurrence, and the incidence of recurrence approaches 20% [7], resulting in distant progression. Similar results were reported by the GOG-258 trial [5]. Compared with pelvic RT, CTx alone consistently shows a higher incidence of severe toxicity, including mortality [5, 7]. Therefore, the adjuvant treatment choice should be determined by understanding the different roles of RT and CTx and by considering treatment-related toxicity, which greatly affects patient quality of life. Notably, the role of adjuvant RT for serous and clear cell adenocarcinoma needs to be discussed between GYNs and ROs because the divergence in opinion between them was noticeable. A subgroup analysis of the PORTEC-3 trial showed that the local relapse rate in patients with serous carcinoma was as high as that in those with endometrioid adenocarcinoma [4]. A collaborative multi-institutional study is warranted to assess the efficacy of RT and CTx in these patient populations.

The results of phase-III randomized controlled trials can largely affect physicians’ pattern of practice, but these effects often occur gradually. Because no study has assessed the pattern of practice since 3 landmark studies (GOG-249, PORTEC-3, and GOG-258 trials), a survey of the up-to-date pattern of practice regarding adjuvant management for EC is timely and necessary. The first limitation of this study would be its relatively low response rate, particularly among GYNs; however, the response rate should be interpreted with caution. As described in Table 2, the number of institutions was high (i.e., >30 in each group) relative to the number of responders. Generally, more than 2 GYNs were working in an academic hospital; some responders responded to the survey as representatives of GYNs in their institution. Therefore, most responders were active members of KGOG, and the actual response rate would be much higher. The second limitation was that this study was based on a cross-sectional survey; no comparison of the use of RT pre- vs. post-publication of some recent clinical studies could be made. However, this limitation was compensated for by mentioning the relevant clinical trials for each question so that the respondents were aware of the clinical trial in question before responding. This guidance may or may not have influenced the responses; however, it was intended to allow the participants to respond in regard to a future patient. The third limitation was that, considering the outcomes of recent clinical trials, we did not pursue some important factors, including the location of the tumor within the uterus or the type and extent of LN dissection (sentinel LN sampling vs. LN dissection), which may be important factors determining the need for EBRT. Performing LN dissection is another important issue to be investigated [18].

Several points should be considered when interpreting the results of our study. In Korea, the overall utility of RT has continued to rise from 24.3% to 29.1% for all cancers [19]. Of the 90 hospitals operating a facility mainly for EBRT, 31 (34.4%) are equipped for brachytherapy [19]. This low rate of brachytherapy facilities is related to the extremely low medical fee for brachytherapy in Korea [20], whereas the high rate of IMRT since 2015 is related to the inclusion of National Health Insurance coverage in Korea. Changes in the medical and economic environments may influence physicians’ preferences; however, physicians must be conscious of what their counterparts (GYNs vs. ROs) achieve in terms of technical development. Surges in new drug investigations or multi-national clinical trials may also affect daily practice for gynecologic cancers in Korea. Nevertheless, opinions vary among nations. For instance, in a survey conducted among the members of the 4 East Asian GOGs in 2017, CTx was the most preferred treatment for locally advanced diseases in Japan, whereas CCRT was preferred in the other countries [21]. Overall, the pattern of practice results may need to be understood within the medical environment and physicians’ societies, which differ from country to country [22].

In conclusion, there are some discrepancies between GYNs and ROs concerning the interpretations and pattern of practice regarding recent clinical trials associated with postoperative management of EC. Although several large-scale clinical trials have been conducted, they lacked appropriate risk assessment; thus, their conclusions were (at best) provisional and should be adopted on a case-by-case basis, depending on local surgical and radiation oncology expertise. Continuous research efforts for optimal management and active communication among experts are essential in this field.

Table S1

Survey questions distributed to the Korean Gynecologic Oncology Group-Korean Radiation Oncology Group members

Table S2

Detailed patient information for specific case questions