Final Analysis of Diagnostic Endoscopic Resection Followed by Selective Chemoradiotherapy for Stage I Esophageal Cancer: JCOG0508

Esophageal cancer is the sixth leading cause of cancer mortality worldwide, and survival is dismal.1Sung H. et al.CA Cancer J Clin. 2021; 71: 209-249Crossref PubMed Scopus (26505) Google Scholar The 5-year overall survival (OS) proportion ranged from 10% to 30% in most countries.2Allemani C. et al.Lancet. 2018; 391: 1023-1075Abstract Full Text Full Text PDF PubMed Scopus (2245) Google Scholar Thus, improving its prognosis is an urgent issue to be addressed. Globally, 85% of esophageal cancer was squamous cell carcinoma (ESCC).3Morgan E. et al.Gastroenterology. 2022; 163: 649-658Abstract Full Text Full Text PDF PubMed Scopus (20) Google Scholar Advances in endoscopic diagnostic technology have increased the detection of early-stage (clinical stage I, T1a–bN0M0) ESCC. Although surgical resection has been the standard of care for stage I ESCC patients,4National Comprehensive Cancer Network 2022https://www.nccn.org/guidelines/guidelines-detail?category=1&id=1433Google Scholar they potentially could be treated by endoscopic resection (ER), a minimally invasive nonsurgical treatment. However, submucosal (T1b) ESCC has a considerable risk of lymph node (LN) metastasis, ranging from 20% to 40%.5Kodama M. et al.Surgery. 1998; 123: 432-439Abstract Full Text Full Text PDF PubMed Scopus (382) Google Scholar,6Tachimori Y. et al.Esophagus. 2013; 11: 21-47Crossref Scopus (44) Google Scholar In addition, even mucosal cancer (T1a) has a potential risk of LN metastasis when pathologic lymphovascular invasion (LVI) exists.7Eguchi T. et al.Mod Pathol. 2006; 19: 475-480Abstract Full Text Full Text PDF PubMed Scopus (244) Google Scholar,8Kim D.U. et al.J Gastroenterol Hepatol. 2008; 23: 619-625Crossref PubMed Scopus (74) Google Scholar Therefore, for patients with stage I ESCC, surgical resection is usually indicated with the 5-year survival proportion of 70% to 80%,6Tachimori Y. et al.Esophagus. 2013; 11: 21-47Crossref Scopus (44) Google Scholar but with organ loss and other complications associated with esophagectomy. To develop more precise treatment with curative and minimally invasive intent, we investigated a new treatment strategy in which diagnostic ER and selective chemoradiotherapy (CRT) based on pathologic tumor depth and LVI for clinical stage I ESCC were used.9Minashi K. et al.Gastroenterology. 2019; 157: 382-390Abstract Full Text Full Text PDF PubMed Scopus (99) Google Scholar Here, we showed the 5-year survival and recurrence patterns of this treatment strategy. The study design is illustrated in Figure 1A. After registration, patients underwent diagnostic ER. Based on the histologic evaluation, additional treatment was selected: those in group A, (pT1a with negative resection margins and no LVI) were followed up without additional treatment; for those in group B, prophylactic CRT was indicated for pT1b with negative resection margins or pT1a with LVI; and for those in group C, definitive CRT was indicated for positive vertical resection margins or uncertain margins for determining cancer-free status. The primary outcome was to evaluate whether the survival of group B and all enrolled patients was equivalent or superior to surgical resection at 5 years. Secondary outcomes were to investigate the time to regional LN failure after CRT, the relationship between failure site and irradiated target volume, and clinicopathologic findings for better or worse survival. Between December 2006 and July 2012, 177 patients were enrolled from 23 institutions in Japan, and 1 patient withdrew consent before treatment. Among the 176 patients who underwent diagnostic ER, 74, 87, and 15 patients were categorized into groups A, B, and C, respectively.9Minashi K. et al.Gastroenterology. 2019; 157: 382-390Abstract Full Text Full Text PDF PubMed Scopus (99) Google Scholar Their clinical courses are shown in Figure 1B (median follow-up period, 81.5 months; range, 0.53–126.1 months). The 5-year progression-free survival (PFS) proportions of group B and all enrolled patients were 86.2% (95% confidence interval [CI], 77.0–91.9) and 87.5% (95% CI, 81.6–91.6), respectively (Figure 1C and D). The 5-year OSs of group B and all enrolled patients were 89.7% (95% CI, 81.1–94.5) and 90.9% (95% CI, 85.6–94.3), respectively (Figure 1E and F). CRT after diagnostic ER (group B) was indicated with the expectation of regional LN control. Among 96 patients (83 in group B and 13 in group C) who received CRT, 11 (9 in group B and 2 in group C) experienced regional LN failure as the first recurrence site (11.5%). The median time to failure after registration was 16.4 months (range, 7.6–53.6 months). The clinicopathologic findings of primary lesions were as follows: Ut/Mt/Lt: 0/9/2; pM3/pSM1/pSM2: 1/2/8; ly0/1: 3/8; and v0/1: 5/6. Regarding the relationship between the failure site and target volume, 8 patients had failure within the irradiated target volume (in target), none recurred outside the irradiated target volume (out of target), 2 patients had failure both in target and out of target, and 1 had failure outside but close to the irradiated target volume (marginal). All out-of-target failures occurred in the abdominal para-aortic LN area. Of the 11 patients, 7 had only regional LN metastases, 6 underwent salvage surgery, and 4 received chemotherapy or radiation therapy. Nine of them died of ESCC, and 1 died of other causes during the study period up to the cutoff date. Among the patients with regional LN metastasis as the first recurrence site, most patients underwent salvage surgery and/or chemotherapy. However, only 1 survived until the final follow-up. This indicates that salvage treatment after recurrence may be ineffective. Therefore, to improve survival, it is most important to reduce recurrences as much as possible and to identify the risk factor for poor prognosis. Univariable analysis of PFS in group B showed that the patients without vascular invasion or those who received 2 courses of chemotherapy had significantly better PFS (Supplementary Table 1). In contrast, pT1b tumors with LVI showed a trend toward worse PFS (Supplementary Table 2). Then, if LVI is positive or a patient receives insufficient chemotherapy, surgery might be better after ER. The present study has several limitations. First, this study was not a randomized controlled trial, and the survival outcomes were not directly compared with those of surgery. Because of the difference in invasiveness between surgery and ER plus CRT, it might be difficult to get patients’ informed consent for randomization. Another limitation is that this study included only the clinical T1b lesions that can be endoscopically resected, whereas surgical studies included all clinical T1b lesions. To evaluate the efficacy of this treatment strategy, the expected threshold values of OS in this study were set higher than those in the surgical data. In this final analysis at the 5-year follow-up, the 5-year OS was still higher than the expected and threshold values, both in patients who received prophylactic CRT (primary endpoint) and in all enrolled patients (key secondary endpoint). We also did not conduct a quality-of-life survey, because this had been not widely performed at the time. When planning future clinical trials, it is necessary to include an analysis of patient outcomes rather than survival alone. In conclusion, the 5-year survival data of patients who underwent diagnostic ER followed by selective CRT were comparable to surgical outcome of clinical stage I ESCC. However, in-target regional LN failure was observed in 11.5% of patients who received CRT. Although further investigation should be performed to improve the efficacy of this treatment strategy for clinical stage I ESCC, this strategy should be considered standard for minimally invasive treatment. JCOG-GIESG Investigators: Junki Mizuaswa, ME1; Kohei Takizawa, MD, PhD2; Ikuo Aoyama, MD3; Akiyoshi Ishiyama, MD4; Noboru Kawata, MD5; Daisuke Kikuchi, MD, PhD6; Noboru Hanaoka, MD7; Ichiro Oda, MD8; Yoshinori Morita, MD, PhD9; Masahiro Tajika, MD, PhD10; Junko Fujiwara, MD11; Yoshinobu Yamamoto, MD, PhD12, Chikatoshi Katada, MD, PhD13; Shinichiro Hori, MD, PhD14; Hisashi Doyama, MD, PhD15; Tsuneo Oyama, MD, PhD16; Hiroko Nebiki, MD, PhD17; Kenji Amagai, MD18; Yutaro Kubota, MD, PhD19; Yasuhiro Inokuchi, MD, PhD20; Nozomu Kobayashi, MD, PhD21; Takuto Suzuki, MD, PhD22; Kingo Hirasawa, MD23; Toshihisa Takeuchi, MD24; and Tomohiro Kadota, MD, PhD1; from the 1Japan Clinical Oncology Group Data Center/Operations Office, National Cancer Center Hospital, Tokyo, Japan; 2Department of Endoscopy, Koyukai Shin-Sapporo Hospital, Hokkaido, Japan; 3Department of Therapeutic Oncology, Kyoto University Graduate School of Medicine, Kyoto, Japan; 4Department of Upper Gastrointestinal Medicine, Cancer Institute Hospital of Japanese Foundation for Cancer Research, Tokyo, Japan; 5Division of Endoscopy, Shizuoka Cancer Center, Shizuoka, Japan; 6Department of Gastroenterology, Toranomon Hospital, Tokyo, Japan; 7Department of Gastroenterology, Osaka Red Cross Hospital, Osaka, Japan; 8Endoscopy Division, National Cancer Center Hospital, Tokyo, Japan; 9Department of Gastroenterology, Kobe University International Clinical Cancer Research Center, Hyogo, Japan; 10Department of Endoscopy, Aichi Cancer Center, Nagoya, Japan; 11Department of Endoscopy, Tokyo Metropolitan Cancer and Infectious Diseases Center, Komagome Hospital, Tokyo, Japan; 12Department of Gastroenterological Oncology, Hyogo Cancer Center, Hyogo, Japan; 13Department of Gastroenterology, Kitasato University School of Medicine, Kanagawa, Japan; 14Department of Gastroenterology, National Hospital Organization, Shikoku Cancer Center, Matsuyama, Japan; 15Department of Gastroenterology, Ishikawa Prefectural Central Hospital, Ishikawa, Japan; 16Department of Endoscopy, Saku Central Hospital Advanced Care Center, Nagano, Japan; 17Department of Gastroenterology, Osaka City General Hospital, Osaka, Japan; 18Department of Gastroenterology, Ibaraki Prefectural Central Hospital, Ibaraki Cancer Center, Ibaraki, Japan; 19Department of Gastroenterology, Showa University Hospital, Tokyo, Japan; 20Department of Gastroenterology, Kanagawa Cancer Center, Kanagawa, Japan; 21Department of Gastroenterology, Tochigi Cancer Center, Tochigi, Japan; 22Department of Endoscopy, Chiba Cancer Center, Chiba, Japan; 23Department of Gastroenterology, Yokohama City University Medical Center, Yokohama, Japan; and 24Department of Gastroenterology, Osaka Medical and Pharmaceutical University, Osaka. We thank the members of the JCOG Data Center/Operations Office for their support: Hiromi Katsuki and Harumi Kaba for data management; Masashi Wakabayashi and Gakuto Ogawa for statistical analysis; and Drs Yukinori Kurokawa, Kenichi Nakamura, Kozo Kataoka, Atsuo Takashima, and Hiroshi Katayama for study design, protocol development, and medical review in the study management process. Keiji Nihei, MD (Conceptualization: Lead; Data curation: Lead; Formal analysis: Lead; Investigation: Lead; Methodology: Lead; Project administration: Lead; Resources: Lead; Supervision: Lead; Validation: Lead; Visualization: Lead; Writing – original draft: Lead; Writing – review & editing: Lead); Keiko Minashi, MD (Conceptualization: Lead; Data curation: Lead; Formal analysis: Supporting; Investigation: Lead; Methodology: Lead; Project administration: Lead; Resources: Lead; Supervision: Lead; Validation: Lead; Visualization: Lead; Writing – original draft: Lead; Writing – review & editing: Lead); Tomonori Yano, MD, PhD (Conceptualization: Lead; Data curation: Lead; Formal analysis: Equal; Investigation: Lead; Methodology: Lead; Project administration: Lead; Resources: Lead; Supervision: Lead; Writing – original draft: Lead; Writing – review & editing: Lead); Tadakazu Shimoda, MD (Investigation: Lead; Writing – review & editing: Equal); Haruhiko Fukuda, MD (Conceptualization: Lead; Data curation: Lead; Formal analysis: Lead; Funding acquisition: Lead; Investigation: Lead; Methodology: Lead; Project administration: Lead; Supervision: Lead; Validation: Lead; Writing – original draft: Lead); Manabu Muto, MD, PhD (Conceptualization: Lead; Data curation: Lead; Formal analysis: Lead; Funding acquisition: Lead; Investigation: Lead; Methodology: Lead; Project administration: Lead; Resources: Lead; Supervision: Lead; Writing – original draft: Lead); Junki Mizusawa, ME (Data curation: Lead; Formal analysis: Lead; Methodology: Lead; Project administration: Lead; Software: Lead; Supervision: Lead; Validation: Lead; Visualization: Lead; Writing – original draft: Lead; Writing – review & editing: Lead); Kohei Takizawa, MD, PhD (Data curation: Equal; Investigation: Equal; Project administration: Equal; Writing – review & editing: Equal); Ikuo Aoyama, MD (Data curation: Equal; Investigation: Equal; Resources: Equal; Writing – review & editing: Equal); Akiyoshi Ishiyama, MD (Data curation: Equal; Investigation: Equal; Resources: Equal; Writing – review & editing: Equal); Noboru Kawata, MD (Data curation: Equal; Investigation: Equal; Resources: Equal; Writing – review & editing: Equal); Daisuke Kikuchi, MD, PhD (Data curation: Equal; Investigation: Equal; Resources: Equal; Writing – review & editing: Equal); Noboru Hanaoka, MD (Data curation: Equal; Investigation: Equal; Resources: Equal; Writing – review & editing: Equal); Ichiro Oda, MD (Data curation: Equal; Investigation: Equal; Validation: Equal; Writing – review & editing: Equal); Yoshinori Morita, MD, PhD (Data curation: Equal; Investigation: Equal; Resources: Equal; Writing – review & editing: Equal); Masahiro Tajika, MD, PhD (Data curation: Equal; Investigation: Equal; Resources: Equal; Writing – review & editing: Equal); Junko Fujiwara, MD (Data curation: Equal; Investigation: Equal; Resources: Equal; Writing – review & editing: Equal); Yoshinobu Yamamoto, MD, PhD (Data curation: Equal; Investigation: Equal; Resources: Equal; Writing – review & editing: Equal); Chikatoshi Katada, MD, PhD (Data curation: Equal; Investigation: Equal; Resources: Equal; Writing – review & editing: Equal); Shinichiro Hori, MD, PhD (Data curation: Equal; Investigation: Equal; Resources: Equal; Writing – review & editing: Equal); Hisashi Doyama, MD, PhD (Data curation: Equal; Investigation: Equal; Resources: Equal; Writing – review & editing: Equal); Tsuneo Oyama, MD, PhD (Data curation: Equal; Investigation: Equal; Resources: Equal; Writing – review & editing: Equal); Hiroko Nebiki, MD, PhD (Data curation: Equal; Investigation: Equal; Resources: Equal; Writing – review & editing: Equal); Kenji Amagai, MD (Data curation: Equal; Investigation: Equal; Resources: Equal; Writing – review & editing: Equal); Yutaro Kubota, MD, PhD (Data curation: Equal; Investigation: Equal; Resources: Equal; Writing – review & editing: Equal); Yasuhiro Inokuchi, MD, PhD (Data curation: Equal; Investigation: Equal; Resources: Equal; Writing – review & editing: Equal); Nozomu Kobayashi, MD, PhD (Data curation: Equal; Investigation: Equal; Resources: Equal; Writing – review & editing: Equal); Takuto Suzuki, MD, PhD (Data curation: Equal; Investigation: Equal; Resources: Equal; Writing – review & editing: Equal); Kingo Hirasawa, MD (Data curation: Equal; Investigation: Equal; Resources: Equal; Writing – review & editing: Equal); Toshihisa Takeuchi, MD (Data curation: Equal; Investigation: Equal; Resources: Equal; Writing – review & editing: Equal); Tomohiro Kadota, MD, PhD (Data curation: Equal; Formal analysis: Equal; Investigation: Equal; Validation: Equal; Writing – review & editing: Equal). This was a multi-institutional, single-arm prospective confirmatory phase III study conducted in accordance with the Declaration of Helsinki and the Japanese Ethical Guidelines for Clinical Studies Involving Human Subjects. The institutional review boards of all participating hospitals approved the study protocol. This trial was registered with the UMIN Clinical Trials Registry (UMIN000000553). The cutoff date for data collection was August 14, 2017. The eligibility criteria and study overview were described previously.1Minashi K. et al.Gastroenterology. 2019; 157: 382-390Abstract Full Text Full Text PDF PubMed Scopus (101) Google Scholar The surgeon and physician also fully explained that the initial standard treatment for submucosal esophageal cancer is surgical resection or chemoradiation, and even more so in the case of noncurative resection by ER. Patients who agreed after full informed consent were enrolled. CRT was initiated between 29 and 70 days after ER, and the chemotherapy regimen comprised 2 courses of continuous intravenous administration of 5-fluorouracil (700 mg/m2/day, days 1–4 and 29–32) and bolus injection of cisplatin (70 mg/m2/day, days 1 and 29). The chemotherapy regimen was based on previous clinical trials for stage I ESCC conducted in JCOG9708.2Kato H. et al.Jpn J Clin Oncol. 2009; 39: 638-643Crossref PubMed Scopus (223) Google Scholar Chemotherapy and radiotherapy were initiated simultaneously on day 1. There was no scheduled intermission for radiotherapy. The dosage of radiotherapy was 41.4 Gy/23 fractions (fr) over 5 weeks for prophylactic CRT (group B) and 50.4 Gy/28 fr over 6 weeks for definitive CRT (group C), delivered with megavoltage equipment (≥6 MV). Computed tomography (CT)–based 3-dimensional treatment planning was required for all patients treated by CRT. The area of prophylactic irradiation covered regional LNs because of the risk of LN metastasis. Among these, supraclavicular, upper mediastinal, and subcarinal LNs were irradiated in patients with primary lesions located in the upper thoracic esophagus. Mediastinal and perigastric LNs were included in the field of irradiation when treating tumors in the middle or lower thoracic esophagus, whereas celiac LNs were included in treating primary tumors in the lower esophagus. In definitive CRT (group C), after the total dose of 41.4 Gy was delivered to regional LNs, a boost dose of 9 Gy was administered to the primary site because of a positive margin after ER. To avoid excessive dose delivery to the heart, a multiple-field technique was required in cases of irradiation to the middle or lower mediastinal LNs. Five-year OS and PFS were evaluated in group B and all registered patients. The definition of endpoints and follow-up details were also described in the previous report.1Minashi K. et al.Gastroenterology. 2019; 157: 382-390Abstract Full Text Full Text PDF PubMed Scopus (101) Google Scholar Follow-up to determine patient survival and disease recurrence continued at least every 6 months until all enrolled patients had been followed for 5 years. After the final follow-up, patients who experienced regional LN failures as the first recurrent site were extracted from groups B and C that received CRT. LN failures that occurred outside the regional area were defined as distant metastases. Regional LN failures that occurred simultaneously with distant metastases were included in this investigation. We examined the time to regional LN failures, survivals, and the clinicopathologic findings of the primary tumor among these patients. In addition to radiation therapy planning data, including target volumes and dose distributions, which had been collected for radiation therapy quality assurance, diagnostic CT images, which confirmed regional LN failures, were collected for central review. For each patient, the regional LN failure site was verified, and the relationship between the failure site and the target volume and dose distribution of elective nodal irradiation was evaluated. All relevant images and materials were centrally reviewed by 2 physicians in charge of this trial (KN and KM). The relationship between failure sites and irradiated target volumes was classified into 3 patterns: in-target, out-of-target, and marginal failures. Marginal failure was defined as LN failure outside but close to the target volume, receiving insufficient dosage of about 50% of the prescription dose. This trial was originally designed to investigate whether the 3-year OS of patients in group B was comparable to or better than that of surgery as the primary endpoint. The 3-year OS of all enrolled patients was evaluated as the key secondary endpoint. As initially planned, the follow-up was extended by an additional 2 years after the primary analysis. Clinically relevant 5-year OS and PFS of group B and all enrolled patients were evaluated in this final analysis. The other details of the statistical analyses, such as sample size calculation and endpoint estimation, were described in the previous report.1Minashi K. et al.Gastroenterology. 2019; 157: 382-390Abstract Full Text Full Text PDF PubMed Scopus (101) Google Scholar All statistical analyses were conducted using SAS 9.4 (SAS Institute Inc).Supplementary Table 1Univariable Analysis of 5-Year PFS for 83 Patients Who Received Prophylactic Chemoradiotherapy in Group BVariablesn5-year PFS, % (95% CI)HR (95% CI)Lymphatic invasion –4689.1 (75.8–95.3)1.447 (0.486–4.306)P = .5044 +3783.7 (67.3–92.3)Vascular invasion –5492.6 (81.5–97.2)3.149 (1.029–9.638)P = .0339 +2975.7 (55.7–87.6)Tumor depth pM3/pSM12993.1 (75.1–98.2)1.819 (0.500–6.612)P = .3567 pSM25483.3 (70.3–90.9)Number of chemotherapy courses 11369.2 (37.3–87.2)0.255 (0.083–0.782)P = .0099 27090.0 (80.2–95.1)NOTE. Patients who had no vascular invasion or those who received 2 courses of chemotherapy had significantly better PFS.HR, hazard ratio; M, mucosal; SM, submucosal. Open table in a new tab Supplementary Table 2Univariable Analysis of 5-Year PFS According to the Depth and LVI of the Primary Lesions in Group BDepthLVI5-year PFS, % (95% CI)HR (95% CI)T1a (M)LVI (+) (n = 14)92.9 (59.1–99.0)1T1b (SM1)LVI (−) (n = 8)100 (100–100)Not estimatedT1b (SM1)LVI (+) (n = 7)85.7 (33.4–97.9)1.05 (0.10–11.58)T1b (SM2)LVI (−) (n = 20)90.0 (65.6–97.4)0.71 (0.10–5.04)T1b (SM2)LVI (+) (n = 34)79.3 (61.4–89.6)1.71 (0.36–8.04)NOTE. Classified by tumor depth and LVI, pT1b tumors with LVI showed a trend toward worse PFS.HR, hazard ratio; M, mucosal; SM, submucosal. Open table in a new tab NOTE. Patients who had no vascular invasion or those who received 2 courses of chemotherapy had significantly better PFS. HR, hazard ratio; M, mucosal; SM, submucosal. NOTE. Classified by tumor depth and LVI, pT1b tumors with LVI showed a trend toward worse PFS. HR, hazard ratio; M, mucosal; SM, submucosal.