Abstract
Background/Aim: The aim of this multi-center retrospective study was to investigate the results of stereotactic body radiotherapy (SBRT) for pulmonary oligometastases from esophageal cancer. Patients and Methods: Oligometastases from the esophagus were identified from a dataset of a Japanese survey. The Kaplan–Meier method and Cox regression were applied to perform analyses. Results: A total of 114 patients with 132 pulmonary oligometastases were collected. The 3-year local control rate, freedom from further metastasis (FFFM) rate and overall survival (OS) rate were 70.2%, 25.3% and 37.5%, respectively. Performance status (PS) (PS 1 vs. PS 0, p<0.01), disease-free interval (p=0.03) and history of local therapy for metastasis (p=0.01) had significant relationships with FFFM and only PS was an independent prognostic factor for OS (PS 1 vs. PS 0, p=0.02; PS 2-3 vs. PS 0, p=0.04). Conclusion: SBRT for pulmonary oligometastases from esophageal cancer provided good local control and survival.
- Stereotactic body radiotherapy (SBRT)
- oligometastases
- oligo-recurrence
- esophageal cancer
- prognostic factor
Surgical resection for pulmonary metastases has been performed for several decades and the key to success has been the selection criterion of only patients in whom all known disease could be removed and the primary site had been fully controlled (1). Thereafter, the concept of oligometastases, which is a limited number of metastatic lesions in a single organ or a few organs, was proposed and was shown to be a good candidate for local therapy (2). This concept of oligometastases has since been extended according to the control of the primary site to include oligo-recurrence, sync-oligometastases and unclassified oligometastases (3-5).
Radiotherapy, another important method for the treatment of cancer, provides a broader access to tumors for local treatment. There have been some reports about stereotactic body radiotherapy (SBRT) for pulmonary metastases (5-9). Recently, the results of a Japanese nationwide survey on SBRT for pulmonary oligometastases have been published (10). The results showed a significantly poor overall survival in patients with oligometastases from the esophagus (Figure 1). Since metastatic esophageal cancer has been regarded as a systemic disease, patients with good or moderate performance status (PS) were typically treated by systemic therapy and/or palliative/best supportive care (11). However, the survival of patients who received chemotherapy for metastatic esophageal cancer has also been poor (12). The outcomes of SBRT, which is an intensive local therapy, for pulmonary oligometastases from the esophagus might be an effective method and should therefore be investigated. The large database from the nationwide survey would enable investigation of large subsets for analyses. The aim of this study was to determine the outcomes of SBRT for pulmonary oligometastases from the esophagus and to identify factors affecting survival.
Overall survival after SBRT for pulmonary oligometastases according to esophagus origin or other origins. There was a significant survival difference in the log-rank test between patients with oligometastases originating from the esophagus and with oligometastases originating from other sites (p<0.01) (a). The overall survival analyses (primary endpoint) were previously reported (10). Overall survival of esophageal cancer after initial treatment (b).
Patients and Methods
Inclusion criteria. The inclusion criteria for the nationwide survey were that the number of pulmonary metastases was limited to 1-5, the primary site and metastatic lesions in other organs were controlled before the commencement of SBRT, SBRT was performed between January 2004 and June 2015, the biological effective dose (BED10) of SBRT was 75 Gy or more and the dose per fraction was 4 Gy or more. BED10 was calculated by the following formula: BED=nd [1+d/(α/β)], where n is the number of fractions, d is dose per fraction and α/β ratio is applied at 10 Gy for the tumors. Cases that included the word “(o)esophagus” or “(o)esophageal” in the primary site column were extracted from the dataset.
Data collection, definitions of terms and patients' consent. This survey was a retrospective, observational, multicenter study in Japan. When the survey was performed, definitions of terms were sent to the collaborators of each institute. Disease-free interval (DFI) was defined as the interval between the date when the primary site was controlled and the date when the first metastasis was confirmed. The starting day of DFI was the day of surgery or the last day of radiotherapy. Adjuvant chemotherapy in the initial treatment was not considered. In cases of simultaneous existence of metastasis with the primary lesion, DFI was regarded as 0. Oligo-recurrences, sync-oligometastases and unclassified oligometastases were defined as DFIs of ≥6 months, 0 months and 0-6 months, respectively. The number of metastases was counted at the time of emergence of the SBRT-targeting tumor. Local failure was defined as enlargement of the irradiated tumor, and local control (LC) was defined as freedom from local failure. Patients were divided into the LC group and the local failure group according to the status of the irradiated tumor. Freedom from further metastasis (FFFM) was defined as freedom from any recurrence or metastasis excluding enlargement of the irradiated tumor and recurrence of the primary site. Toxicity of SBRT was reported according to the National Cancer Institute Common Terminology Criteria for Adverse Events version 4.0 translated by the Japanese clinical oncology group (CTCAE-JCOG). This study was approved by the ethical committee of a senior facility (Ethics Committee of Toho University Omori Medical Center, reference number: 27-148). Informed consent was waived due to the retrospective nature of this study. All participating institutions were guaranteed the chance to opt out of participation in this study by giving the information of this study via the Internet or posters, and opt-out consent was obtained from all patients.
Analysis. Cumulative LC rate, FFFM rate and overall survival (OS) rate were calculated using Kaplan–Meier curves, and the log-rank test was used to compare the curves. The time to an event was calculated from the first day of SBRT to the day that an event was confirmed. The Kaplan–Meier OS curve was also described from the first day of DFI (i.e., completion of initial treatment). The Cox proportional hazards model was used to perform univariate analyses and multivariate analyses (MVA). Continuous variables were divided according to the median into categorical variables. In MVA, a stepwise backward elimination/forward addition approach using the Akaike information criterion (AIC) was used. A p-value of less than 0.05 was defined as significant. EZR version 1.37 (Saitama Medical Center, Jichi Medical University, Saitama, Japan), a modified version of R commander (R Foundation for Statistical Computing, Vienna, Austria), was used for analyses (13).
Patients' demographics at the time of SBRT.
Results
Treatment outcomes. A total of 114 patients were enrolled from 28 institutions. One-hundred forty-six pulmonary oligometastatic tumors from esophageal cancer were identified at the time of emergence of the SBRT-targeting tumor: 132 oligometastatic tumors were treated with SBRT and 14 oligometastatic tumors were controlled by surgical resection and so on. Patient demographics at the time of SBRT are shown in Table I. All of the patients completed the planned SBRT, and tumor and treatment characteristics are summarized in Table II. The median follow-up period from SBRT for all patients was 17.8 months (range=0.2-112.2 months) and that from initial treatment was 45.1 months (range=2.2-153.8 months). The 1-, 3-, and 5-year LC rates of SBRT-treated tumors were 89.9% [95% confidence interval (95% CI)=82.0-94.5%], 81.4% (95%CI=70.2-88.7%), and 81.4% (95%CI=70.2-88.7%), respectively (Figure 2). The 1-, 3-, and 5-year FFFM rates after SBRT were 43.9% (95%CI=33.5-53.7%), 25.3% (95%CI=15.7-35.9%) and 22.7% (95%CI=13.3-33.7%), respectively, and the median FFFM period was 10.8 months (95%CI=8.7-13.4 months) (Figure 2). There was no significant difference in FFFM between the LC group and the local failure group (p=0.42). Patterns of reported first failure site(s) after SBRT were local failure in 9 patients, recurrence of the primary site in 2, lymph node metastases in 17, further lung metastases or pleural dissemination in 23, liver metastases in 2, brain metastases in 2, others in 7 and unknown in 19. The 1-, 3-, and 5-year OS rates from SBRT were 80.4% (95% CI: 71.5-86.8%), 37.5% (95%CI=27.0-48.0%) and 23.0% (95%CI=13.5-34.0%), respectively, and the median survival time (MST) was 27.1 months (95%CI=20.3-33.2 months) (Figure 1). Of the 114 patients, 50 died of esophageal cancer and 16 died of other causes. The 1-, 3-, and 5-year OS rates from initial treatment were 98.1% (95%CI=92.6-99.5%), 74.2% (95%CI=64.1-81.8%) and 42.6% (95%CI=31.4-53.3%), respectively, and the MST was 57.2 months (95%CI=49.9-60.7 months). Adverse events after SBRT were assessed in 90 of the 114 patients. Radiation pneumonitis of grade 2 or worse occurred in 6 patients, including 1 patient with grade 3 and 1 patient with grade 5. Radiation dermatitis of grade 2 and grade 3 occurred in 3 patients and 1 patient, respectively, and a gastrointestinal adverse event of grade 4 occurred in 1 patient.
Tumor and treatment characteristics.
Local control rates of SBRT sites (a). Freedom from further metastasis periods after SBRT (b) and freedom from further metastasis according to local control, with no significant difference (p=0.42) (c).
Analyses of prognostic factors and predictive factors. The results of univariate Cox regression analyses for LC, FFFM and OS are shown in Table III. PS (PS 1 vs. PS 0, p=0.03), oligometastatic state (sync-oligometastases vs. oligo-recurrence, p=0.04) and DFI (≥13 months vs. <13 months, p=0.01) showed significant relationships with FFFM. PS (PS 1 vs. PS 0, p=0.01; PS 2-3 vs. PS 0, p<0.01) and maximum tumor diameter (≥1.3 cm vs. <1.3 cm, p=0.03) were significant prognostic factors for OS, and there was no significant predictive factor for LC. The results of multivariate Cox regression analyses are shown in Table IV. Significant prognostic factors for FFFM were PS [PS 1 vs. PS 0, hazard ratio (HR)=2.10, 95%CI=1.20-3.66, p<0.01), DFI (≥13 months vs. <13 months, HR=0.54, 95%CI=0.30-0.95, p=0.03) and local therapy history for metastases (Yes vs. No, HR=2.22, 95%CI=1.18-4.17, p=0.01). Only PS was an independent prognostic factor for OS (PS 1 vs. PS 0, HR=1.84, 95%CI=1.06-3.18, p=0.02; PS 2-3 vs. PS 0, HR=2.47, 95%CI=1.03-5.87, p=0.04). Multivariate analysis for LC was not performed because no potential factors emerged from the results of univariate analyses.
Results of univariate Cox proportional analyses.
Results of multivariate Cox proportional analyses using a stepwise selection.
Discussion
This study was performed as subset analyses of a Japanese multicenter survey of SBRT for pulmonary oligometastases (10). Pulmonary oligometastases that have been treated by SBRT include oligometastases from various primary sites, and the effectiveness of SBRT for oligometastases has been compared according to the primary sites (8, 9, 14, 15). Primary endpoint analyses of SBRT for pulmonary oligometastases in our previous study were also conducted by a comparison according to the primary sites, and it was shown that patients with oligometastases of esophagus origin had significantly poorer survival. These results were the motivation for performing subset analyses of SBRT for esophageal pulmonary oligometastases using results of a multicenter survey in Japan, and the results of this study are therefore considered to be important and meaningful. Furthermore, to the best of our knowledge, the dataset used for analyses in this study is the largest so far used for analyses of SBRT and this study is the first study on SBRT for esophageal pulmonary oligometastases only (16).
The results of analyses showed that the MST from SBRT was 27.1 months and that the MST from initial treatment was 57.2 months. The MST from SBRT reflected survival from an M1 or rM1 state according to the UICC TNM classification 7th ed. The actual MST from diagnosis of M1 or rM1 would be longer because 31.3% of the patients in this cohort had a history of local therapy for metastases and 36.8% of the patients had received chemotherapy before SBRT. If these results are compared to previous results for esophageal cancer, our cohort showed longer survival. Actually, it has been reported that MSTs of patients with metastatic esophageal cancer were around 5 to 21 months (17-20). Regarding recurrence after curative surgery for esophageal cancer, survival of patients with distant metastasis was worse than that of patients with locoregional recurrence (21, 22). However, due to our inclusion criteria, good survival results were obtained even for patients with distant metastasis, and the criteria would be useful, but with caution, in clinical practice. Since esophageal cancer has sometimes been treated by (chemo)radiation, a considerable number of patients in this study had previously received radiotherapy for a primary lesion and careful attention should therefore be paid to SBRT (Table I). From another point of view, safety of SBRT for pulmonary oligometastases was demonstrated even in such patients, considering the toxicity of previous salvage radiotherapy for lymph node oligo-recurrence of esophageal cancer, as shown in a previous study (23). Although the results for patients with esophageal cancer in this study were good considering other reports about esophageal cancer, patients with oligometastases originating from the esophagus showed poor survival compared with oligometastases originating from other sites (Figure 1). One of the reasons for the shorter survival in the esophagus cohort would be the invasiveness of primary lesion treatment. Since the primary site was controlled well in the current study and only two primary failures in the esophagus occurred during follow-up, PS was the only significant prognostic factor of OS in MVA. The significance of PS would be due to various factors such as the state of nutrition, frailty, and body weight. Patients with poor PS might therefore not be good candidates for aggressive treatment.
DFI, history of local therapy and PS showed relationships with FFFM. Since DFI was sometimes reported as a factor related to OS and history of local therapy means a prior diagnosis of rM1, these results are not surprising (8, 24). Although the results of this study showing that DFI and a history of local therapy were significantly related to FFFM are reasonable results, it was shown that neither DFI nor a history of local therapy was a prognostic factor for OS. They would not overcome the effect of PS in OS. In FFFM, the role of PS was under discussion because patients with PS 1 showed a high rate of further metastasis, whereas patients with PS 2-3 tended to have a lower rate of further metastasis. Patients with PS 2-3 who had very short survival might have died of other causes before the development of further metastasis.
Results of further analysis for FFFM unfortunately showed that successful LC by SBRT did not contribute to a decrease in the rate of FFFM, and Kaplan–Meier FFFM curves in the LC group and local failure group were similar (Figure 2). However, the results suggested that there were possible long tails of the curve, which showed long FFFM in both groups. These patients would obtain a benefit from SBRT. If local failure did not occur, some censored patients in the local failure group of FFFM might have obtained longer survival. The benefit of SBRT will be demonstrated in future prospective phase III trials following a recent successful randomized phase II trial (25). However, the trial of oligometastases originating only from the esophagus would be difficult due to few candidates.
There are several limitations in the current study. First, this study was a retrospective study and the number of esophageal pulmonary oligometastases was relatively small. Therefore, there were some selection biases and missing data. Second, various treatment protocols were included in the analysis. There were a variety of total doses, methods of dose prescription and fractionation schema at each institution.
In conclusion, patients with oligometastases from the esophagus showed better survival with 27.1 months of MST compared to previously reported outcomes. The key to success might be the inclusion criteria of this study that oligometastases were limited to the lungs and treated with curative intent under the control of the primary site and other extrathoracic oligometastases. Adverse events were tolerable although 39.2% of the patients were treated with (chemo)radiotherapy for control of the primary site. This study suggested that there are some patients who have FFFM for a relatively long time and might obtain the benefit of LC by SBRT.
Acknowledgements
The Authors acknowledge the collaboration of many radiation oncologists in Japan. The Authors are grateful to Drs. Nanae Yamaguchi, Kuniaki Katsui, Takuya Yamazaki, Hideya Yamazaki, Takashi Shintani, Hiroshi Onishi, Atsushi Nishikawa, Akira Anbai, Masatoki Ozaki, Shizuko Ohashi, Nobuhisa Mabuchi, Michiko Imai, Hisashi Kaizu, Keisuke Fujimoto, Katsuyuki Shirai, Kentaro Yamamoto, Kazunari Yamada, Seiji Kubota, Shigetoshi Shimamoto, Takanori Fukuda, Tsunehiko Kan and Yoshihiko Manabe.
Footnotes
Authors' Contributions
TY drafted the protocol, collected patients, performed data analysis and interpretation and wrote the manuscript. NY conceived and designed the study, drafted the protocol and collected patients. YM, YD, R-JO, HY, HK and MA collected patients. KJ advised data analysis and interpretation and revised the manuscript. All Authors read and approved the manuscript.
Conflicts of Interest
The Authors declare no conflicts of interest associated with this manuscript.
Funding
This study was supported by Grant-in-aid for research on radiation oncology of JASTRO 2015-2016.
- Received February 23, 2020.
- Revision received March 1, 2020.
- Accepted March 4, 2020.
- Copyright© 2020, International Institute of Anticancer Research (Dr. George J. Delinasios), All rights reserved







