Adjuvant Chemotherapy for Endometrial Cancer (ACE) trial: A randomized phase II study for advanced endometrial carcinoma

Abstract This study evaluated the feasibility and efficacy of three postoperative adjuvant chemotherapy regimens for endometrial cancer. Endometrioid cancer patients with intermediate‐risk stage I and II or high‐risk stage III and IV disease were randomly assigned to receive six cycles of either paclitaxel‐epirubicin‐carboplatin (TEC), paclitaxel‐anthracycline (doxorubicin)‐carboplatin (TAC), or dose‐dense paclitaxel‐carboplatin (ddTC). The primary end‐point was the completion rate (CRate) of six cycles of treatment. The secondary end‐points were progression‐free survival (PFS) and overall survival (OS). One hundred and one patients were treated as follows: 33 received TEC, 33 TAC, and 35 ddTC. The CRates for TEC, TAC, and ddTC were 94%, 64%, and 69%, respectively (P = .005). The TEC CRate was significantly higher than for the other two groups. However, the PFS and OS outcomes were not statistically different between the three groups. The 2‐year survival rates were 94%, 97%, and 97% for TEC, TAC, and ddTC, respectively. When compared to the current standard treatments for endometrial cancer, TEC is a promising candidate for a phase III trial based on its significantly superior CRate and equivalent PFS and OS. This study is registered with UMIN Clinical Trials Registry (UMIN000008911).


| INTRODUC TI ON
The per capita incidence of endometrial cancer, already the most common gynecologic cancer worldwide, is steadily increasing in developed countries. 1 The early-stage forms of endometrial cancer are usually highly curable by surgical treatment alone, whereas advanced stages require adjuvant interdictions such as radiotherapy and chemotherapy. Platinum and anthracycline drugs have long been used as standard adjuvant chemotherapy drugs for advanced and recurrent endometrial carcinomas. 2,3 In one study, the standard combination adjuvant treatment with AP was found to be more effective as an adjuvant therapy than whole abdominal irradiation (GOG122), 4 whereas no differences in outcomes between AP and radiotherapy were found in two other trials, one Italian, one Japanese. 5,6 Still, in the NSGO/EORTC trial (Nordic Society of Gynecologic Oncology/European Organization for the Research and Treatment of Cancer) and in the pooled results from two other studies, chemotherapy was found to be superior to radiation therapy. 7 Before the introduction of paclitaxel into the GOG177 study, OS was not consistently improved in most drug comparison trials. The combination of paclitaxel, doxorubicin, and cisplatin (160, 45, and 50 mg/m 2 , respectively) was found to be superior to AP (GOG 177).
The response rates were 57% in the TAP arm of the study and 34% in the AP arm (P < .001); however, despite its better response rate, severe adverse events occurred more frequently with TAP, and it often requires G-CSF support. In one study's TAP arm, there were five treatment-related deaths (3.8%), but none in the AP arm. 8 Thus, considering the severe adverse events associated with TAP, it has not been widely used.
To search for an effective yet tolerable drug to be combined with a paclitaxel regimen for endometrial cancer, various other drug types have been studied. A combination of paclitaxel (175 mg/m 2 ) and carboplatin, with an AUC of 5-6 for six cycles, has been the gold standard regimen for another gynecologic tumor, ovarian cancer.
The TC regimen is now also widely used for endometrial cancer. 2 In one study, the response rate to TC for chemonaïve advanced endometrial cancer was 47%. 9 With the addition of epirubicin to TC, TEC therapy has also been reported to be effective for the treatment of advanced and recurrent endometrial cancers. 10,11 Our own retrospective study showed that the TEC response rates for advanced or recurrent endometrial cancer were superior to TC alone (68% response in the TEC group vs 52% in the TC group). 3 In one study, 12 a TC regimen using ddTC was found to be more effective for ovarian cancer than a triweekly regimen of TC, whereas, in the Mito-7 study, ddTC was not superior to triweekly TC. 13 However, the dose of TC used in the Mito-7 study was not the same as in the other trials mentioned. It is thought that the ddTC regimen is more effective than the triweekly regimen because the dose-interval of ddTC is shorter, which may suppress tumor growth, and the total volume of the ddTC regimen is more than the triweekly regimen. The ddTC regimen has recently been reported to be effective for endometrial cancer as well. [14][15][16] As yet another combination therapy based on TC, TAC is a regimen of paclitaxel, doxorubicin, and carboplatin used for ovarian cancer that can also be effective against endometrial cancer. Recently, the endometrial cancer response rate to TAC was reported to be 67%. 17 To evaluate the feasibility of using one or more of these three combination adjuvant regimens for the routine treatment of advanced stage endometrial cancers, we undertook a randomized study comparing the safety and efficacy of TEC, ddTC, and TAC. This three-armed GOGO EM-3 randomized drug study was designed to find the best candidate for a future phase III trial for treatments of advanced or recurrent endometrial cancer. We have called this the ACE trial.

| Patients
Our study was undertaken as an intergroup cooperative carried out between GOGO, KCOG, TGCU, and EGOG. Patient participation eligibility required that the case be histologically diagnosed with primary endometrial carcinoma. The patient should have received at least a hysterectomy, salpingo-oophorectomy, and pelvic lymphadenectomy and there should be no residual tumor, or the residual tumor should be less than 2 cm in size. 64%, and 69%, respectively (P = .005). The TEC CRate was significantly higher than for the other two groups. However, the PFS and OS outcomes were not statistically different between the three groups. The 2-year survival rates were 94%, 97%, and 97% for TEC, TAC, and ddTC, respectively. When compared to the current standard treatments for endometrial cancer, TEC is a promising candidate for a phase III trial based on its significantly superior CRate and equivalent PFS and OS. This study is registered with UMIN Clinical Trials Registry (UMIN000008911).

| Study design
Patients were registered into the study by an investigator, whereupon the patient was randomly allocated by the data center at a 1:1:1 ratio into one of the three study arms. The minimization method, with surgical stage (I or II vs III or IV) as the adjustment factor, was used for randomization assignment. Completion of six cycles of the regimen was defined as completed.
The 3-week interval treatments were delayed when any of the following adverse events were observed: number of neutrophils fell below 1500/mm 3 ; number of platelets fell below 75,000/mm 3 ; levels for both aspartate aminotransferase and alanine aminotransferase rose above 100 U/L; serum creatinine levels rose above 1.5 mg/ dl; the patient's temperature rose above 38°C; or the patient's performance status fell below 3. Treatments were also interrupted or altered for other adverse events, such as a hematology event greater than grade 3 (Common Terminology Criteria for Adverse Events, version 4.0), an allergic reaction greater than grade 1, or if a physician judged it was appropriate to postpone or skip the treatment, in which case the next treatment course was started within 6 weeks of the previous treatment.
The hematologic criteria used for deciding suspension of the ddTC treatments on days 8 and 15 were if numbers of neutrophils fell below 500/mm 3 or platelets fell below 50,000/mm 3 . If hematologic criteria were not adequate, treatment was delayed for 7 days, and the dose of carboplatin was decreased in the next course. If the hematologic criteria did not return to an adequate status during the next 2 weeks of delay, the ddTC treatment was permanently terminated.
The dose of each chemotherapy was decreased in certain cases, as follows: (a) febrile neutropenia occurred, with neutropenia grade 3 or 4 and fever higher than 38.5°C; (b) neutropenia grade 4 for more than 5 days in a row; (c) grade 4 thrombocytopenia, or grade 3 thrombocytopenia with a bleeding tendency, requiring platelet transfusion; (d) grade 2 or higher peripheral neuropathy; and (e) grade 3 or higher nonhematologic adverse events, except for nausea, diarrhea, and alopecia. The details of the dose reduction scheme are described in Table S1. Following the full completion of the treatment arm, the patient was monitored monthly.

| Statistical analysis
The primary end-point of the study was the CRate of six cycles of each assigned treatment arm. As the purpose of this study was to select a treatment to proceed to the next trial phase, the selection design proposed by Sargent and Goldberg 18 was applied. The CRate for the three regimens was expected to range from 70% to 77%, based on prior studies. 11,14,17 A clinically significant difference in CRate was considered to be 15%, so the difference in true CRate was assumed to be 15%. If the result of this study shows that the highest CRate is 1% or more different from the other two groups, we intended to select that treatment group for the next trial phase.
Under these conditions, for correctly selecting the treatment group with the highest Crate, 35 cases per group were required to make the probability 80% or more.
The secondary end-points of this study were PFS and OS curves, which were estimated by the Kaplan-Meier method and compared by the log-rank test. Progression-free survival and OS were defined as the periods from the day of the beginning of chemotherapy to the day of progression of the disease or the death of the patient from the disease was admitted.
Other data were analyzed using the χ 2 test. If there was a significant difference among the three groups, the multiple comparison test was applied, with Bonferroni's correction. The indicated P values in the tables are unadjusted. The adverse event data were divided into grades 0, 1, 2 and grades 3, 4, 5, and then compared.  Figure 1. The remaining 101 patients were randomly allocated to treatment arms as follows: 33 patients were assigned to the TEC group, 33 to the TAC group, and 35 to the ddTC group. We found no differences between the three groups for the patients' individual characteristics, listed in Table 1.

| Six-cycle CRate
The CRates of the TEC, TAC, and ddTC groups were 94% (31/33), 61% (20/33), and 69% (24/35), respectively. The CRate for TEC was the highest of the three groups ( Table 2). For three ddTC cases, treatment was discontinued by the patient's doctor. In one case, the patient had a grade1 allergy to the paclitaxel, in another, the patient had interstitial pneumonia, and in the third case, the patient had grade 2 neuropathy.
The 2-year survival rate was 94%, 97%, and 97% in the TEC, TAC, and ddTC groups, respectively ( Figure 2). According to Kaplan-Meier analysis, the PFS and OS resulting from the three treatments, TEC, TAC, and ddTC, were not statistically different, and this was true regardless of whether they were treated for intermediate-risk (stage I and II) or high-risk (stage III and IV) tumors ( Figure S1).

| Adverse events
Grade 3 and 4 neutropenia were observed more frequently in the TEC and TAC groups than in the ddTC group (67%, 73%, and 14%, TEC vs ddTC, P = .000015; TAC vs ddTC, P < .000001, respectively). Grades 3 and 4 fever were significantly more frequent in the TAC group than in the ddTC group (21% and 0%, respectively, P = .0044). The frequency of using G-CSF was significantly higher in the TAC group than in the ddTC group (36% vs 3%, P = .00046). The frequency of using G-CSF in the TEC group was 21%, which was not significantly higher than for the ddTC group (P = .025; Table S2). The frequency of other adverse events was  Table 4). The frequency of needing to reduce the drug dose was not different in the three groups. The frequency of treatment delay was significantly more frequent in the ddTC group than in the TEC group (36% and 77%, respectively; P = .0012; Table S3).

| DISCUSS ION
The primary end-point of this comparative study of combination adjuvant chemotherapies was the successful CRate of six cycles of treatment. Under our rigid criteria, the TEC group had the highest CRate, i.e., TEC exhibited the least, or the most tolerable, toxicities.
The results of the 2-year follow-up for the two secondary endpoints, OS and PFS, on the other hand, were not considered to be statistically different among the three groups.
The AUC for carboplatin in each regimen was 4 mg/mL·min in TEC and 5 mg·mL/min in TAC and ddTC. This may be one reason why TEC had the highest CRate. The higher dose of carboplatin The CRate for TEC in the current study was 94%, making it appear superior to the CRates of TAC and ddTC (61% and 69%). The CRate for TEC in the current study was superior to the CRate for TC in the JGOG2043 study, and better than the CRate for seven cycles of TC in the GOG0209 study. 19,20 There were no cardiac failures nor treatment-related deaths for the TEC regimen, despite it including the anthracycline drug epirubicin.
The TC regimen is highly familiar to gynecologic oncologists as a treatment for ovarian cancer, so it is easily used in most hospitals. Note: Multiple answers were allowed. Two doctors in the TAC group selected two answers, one selected "progressive disease and a grade 3 adverse event", the other responded "patient withdrawal and a grade 3 adverse event."  From the results of the GOG0209 study, the TC regimen would be presumed to be the first choice for endometrial cancer as well.

TA B L E 3 Reasons for treatment cessation in patients with advanced endometrial carcinoma
Meanwhile, in the GOG0209 study, the OS for women of Japanese descent tended to be better with the TAP regimen than the TC regimen, although the difference was not statistically significant.
There has been one phase II study that found that the ddTC regimen is superior to the TC regimen for advanced endometrial cancer. 15 However, the CRate for ddTC, at 69%, was relatively low. A careful comparison of the ddTC and TC regimens in a large phase III study is now needed.
The PORTEC-3 trial recently found that a combination of adjuvant chemotherapy and radiotherapy was superior to pelvic radiation therapy alone. 21 The regimen used in that study was two cycles of cisplatin (50 mg/m 2 ) given intravenously concurrently with radiotherapy, followed by four cycles of carboplatin (AUC5) and paclitaxel (175 mg/m 2 ). However, six cycles of TC chemotherapy were not inferior to combined adjuvant chemotherapy and radiotherapy (cisplatin [50 mg/m 2 on days 1 and 29], together with volume-directed external-beam radiation therapy, followed by four cycles of TC). 22 There are limitations to our current study. It was mainly designed to compare only the CRate for each arm. The number of patients per arm (33-35) was relatively small and the period of observation (2 years) is so far very short. This is one reason that PFS and OS were not statistically different in these three groups. Many more cases and a much longer observation period will be needed to determine the best regimen for adjuvant chemotherapy for advanced endometrial cancer.
The lack of data in some cells indicates data were analyzed using the chi-square test. If there was a significant difference among the three groups, the multiple comparison test was performed, with Bonferroni's correction.
chemotherapy. Another consideration is the cost of molecularly targeted therapy, which is much higher than traditional chemotherapies, meaning the latter are still much needed for the routine treatment of endometrial cancer. Thus, their optimal regimens must be elucidated, as we have attempted to do with this study.
In conclusion, the six-cycle CRate for advanced endometrial cancer treatment was highest in the TEC group, compared to TAC and ddTC; however, the PFS and OS of each group were the same. Based on its superior CRate, the TEC regimen is the most feasible of the three regimens for promotion forward for a phase III trial for endometrial cancer.

ACK N OWLED G M ENT
We would like to thank Dr. G.S. Buzard for his constructive critique and editing of our manuscript. There was no source of financial support for this study.