Salvage Radiotherapy for Loco-regional Recurrence of Esophageal Cancer Following Surgery
Article information
Abstract
Purpose
There is few evidence regarding the optimal salvage treatment options for loco-reginal recurrence of esophageal cancer. This study aimed to evaluate the clinical outcomes of salvage radiotherapy (RT) in patients with loco-regional recurrence (LRR) after surgery for esophageal cancer.
Materials and Methods
We retrospectively reviewed 147 esophageal cancer patients who received salvage RT for loco-regional recurrence between 1996 and December 2019. A total dose of 60 Gy in 20 fractions was used for RT alone and 60-70 Gy in 30-35 fractions for concurrent chemoradiotherapy (CCRT).
Results
The patients’ median age was 65 years (range, 41 to 86 years). The median disease-free interval was 13.5 months (1.0 to 97.4 months). After a median 18.8 months follow-up, the 2-year overall survival (OS) and progression-free survival (PFS) rates were 38.1% and 25.9%, respectively. The median OS and PFS were 18.8 and 8.4 months, respectively. The CCRT could not improve OS compared to RT (p=0.336), but there was a trend of better PFS in the CCRT group. Regarding toxicities, the rate of grade 3 or higher toxicity was 10.9% occurring in 16 patients, and it was higher in patients who received CCRT than in the RT alone group (19.6% vs. 6.3%, p=0.023).
Conclusion
Salvage RT alone as well as CCRT could be effective in patients with locoregionally recurrent esophageal cancer.
Introduction
Surgical resection is the most important treatment option for resectable esophageal cancer patients. Following the initial surgery, roughly about half of the patients (42.5%-52.4%), however, experience disease recurrence with the median time to recurrence of 7.0-12.2 months [1-7]. The prognosis of the patients with post-surgical recurrence is usually dismal, whose median survival time is 6.0-8.2 months [4,5]. More favorable clinical outcomes, however, are usually associated in loco-regional recurrence (LRR) setting than in distant metastasis setting [2].
Salvage treatment options applicable for LRR of esophageal cancer include surgery, radiation therapy (RT), and concurrent chemoradiotherapy (CCRT) [8-10]. Among these, salvage CCRT is preferably recommended following previous esophagectomy according to the National Comprehensive Cancer Network (NCCN) guidelines [11], which has been based on several prospective and retrospective studies [9,12-15]. Most studies, however, were single institution-based with limited sample sizes, and were reported mostly from Japan or China [15]. Our institution, as one of the tertiary referral hospitals, has covered a large portion of surgical resection for the esophageal cancer patients in Korea [16], and this study is to report the clinical outcomes, prognostic factors, and toxicity profiles following salvage RT or CCRT for LRR of esophageal squamous cell carcinoma following surgery.
Materials and Methods
1. Patients
We retrospectively reviewed the medical records of the patients who underwent salvage RT-based treatment for LRR of esophageal squamous cell carcinoma at the authors’ institution between January 1996 and December 2019. We excluded the patients (1) who had adenocarcinoma histology, (2) who underwent palliative, non-curative, surgery, (3) who had combined distant metastasis, (4) who were treated with other malignancies within two years, and (5) who received RT dose less than 30 Gy or abandoned the recommended RT course, respectively (S1 Fig.).
Local recurrence was defined as recurrence at the anastomosis site or residual esophagus, and regional recurrence was as recurrence in the regional lymph nodes of esophageal cancer as described in the American Joint Cancer Committee (AJCC) staging manual, respectively. LRR encompassed both local and regional recurrences. At the time of recurrence, disease status was assessed using chest computed tomography (CT) (with or without abdomen pelvis CT) and fluorodeoxyglucose (FDG)-positron emission tomography (PET)/CT. Esophagogastroduodenoscopy (EGD) was recommended if local recurrence was suspected and biopsy confirmation was tried whenever feasible. While chest CT scans were performed on all patients, FDG-PET/CT was conducted in 124 patients (84.3%). EGD was done in 65 (44.2%) and biopsy confirmation was done in 50 (34.0%), respectively. The decision on applying either RT or CCRT, instead of surgical resection, was made through a multidisciplinary discussion considering the patient’s medical condition, technical difficulty, and estimated risk following salvage surgery. The first response evaluation was done in 1 month of RT completion, and the second one was done in 3-4 months thereafter typically with PET-CT. Further follow-up evaluations were scheduled regularly most commonly with chest CT (with or without abdomen pelvis CT) at 3-4 months’ interval for the first 2 years and at 6-12 months’ interval thereafter.
2. Treatment
RT was delivered using 4-10 MV photon beams generated by the linear accelerators (Varian Medical System, Palo Alto, CA). After delineating the gross tumor volume (GTV) based on the available clinical information and imaging studies, the clinical target volume was delineated to cover the GTV with a 5-10 mm margin in all directions. Until 2016, 3-dimensional conformal RT (3D-CRT) using 3-4 beams was mainly used, and intensity-modulated RT (IMRT) has been used more frequently since after 2015, when the Korean National Health Insurance Plan began to cover this technique. In salvage RT alone, 60 Gy in 20 fractions was the typical dose schedule, while 60-70 Gy in 30-35 fractions was preferred in CCRT setting, respectively. For CCRT, 2 cycles of 5-fluorouracil plus cisplatin at 3-weeks’ interval were most commonly employed. Additional systemic chemotherapy after completion of salvage RT or CCRT was offered to nine patients, based on the discretion of the medical oncologist in charge, and the most commonly used regimen was 2-3 cycles of 5-fluorouracil plus cisplatin at 3-weeks’ interval.
3. Endpoints and statistical analysis
The primary endpoints were overall survival (OS) and progression-free survival (PFS), and the secondary endpoint was failure pattern following salvage RT-based treatment. The observation durations were calculated from the dates of RT start until the dates of event, death, or censoring. The rates of OS and PFS were calculated using the Kaplan-Meier method and compared using the log-rank test. The factors with a probability value of < 0.2 were entered into the Cox proportional hazard regression analysis, and a p-value less than 0.05 was considered statistically significant. The acute toxicities were defined as those appearing within three months from the RT start, while the late toxicities were those observed thereafter. The Radiation Therapy Oncology Group toxicity scales and the Common Terminology Criteria for Adverse Events ver. 4.0 were used to evaluate the RT-related toxicities. The proportions of toxicities were compared by the kai square test. All the statistical analyses were performed using the SPSS ver. 27.0 (IBM Corp., Armonk, NY).
Results
1. Patients’ characteristics
The characteristics of 147 patients are summarized in Table 1. The median age of all patients was 65 years (range, 41 to 86 years), and vast majority were male (n=138, 93.9%). The most frequent location was the mid-thoracic esophagus (55.8%) and a majority of patients (86.4%) had performance status of Eastern Cooperative Oncology Group (ECOG) 0-1. The pathologic stages following the initial surgery, according to the AJCC 7th edition were pI in 24 patients (16.3%), pII in 53 (36.1%), and pIII in 71 (47.6%), respectively. Fifty-seven patients (38.8%) underwent adjuvant chemotherapy following initial surgery. Among 71 pIII patients, who were candidates for adjuvant chemotherapy according to the current standard, 27 did not receive adjuvant chemotherapy: seven underwent surgery in the 1990s when adjuvant chemotherapy was not routinely recommended; 12 skipped chemotherapy due to poor performance status; six refused chemotherapy; one developed recurrence before starting adjuvant therapy; and one was enrolled in another clinical trial, respectively. The median disease-free interval (DFI), defined as the time from the initial surgery till the start of salvage RT, was 13.5 months (range, 1.0 to 97.4 months). The sites of recurrence were local in 24 patients (16.3%), regional in 107 (72.8%), and both loco-regional in 16 (10.9%), respectively, and 89 (60.5%) had a single site recurrence. Stages at recurrence were rII in 113 patients (87.6%) and rIII in 16 (12.4%), respectively. Salvage RT alone was applied to 96 patients (65.3%), while CCRT was to 51 (34.7%), respectively. Chemotherapy following salvage RT was administered only in nine patients, all of them underwent CCRT. Over three-quarters of patients (n=116, 78.9%) received 3D-CRT and the rest (n=31, 21.1%) did IMRT, respectively. The median RT dose was 60.0 Gy (range, 30.0 to 70.0 Gy), and the median biological effective dose with α/β=10 (BED10) was 78.0 Gy10 (range, 37.5 to 84 Gy10), respectively.
Patients’ characteristics
2. Treatment outcomes
The median follow-up duration from the salvage RT start was 18.8 months (range, 1.3 to 296.8 months). The follow-up period was not significantly different between RT alone and CCRT groups (17.7 and 19.5 months, respectively, p=0.323). During the follow-up, 96 patients (65.3%) experienced disease progression, and 123 (83.7%) died. The median periods of OS and PFS were 18.8 months and 8.4 months, respectively (Fig. 1A and B). The 2- and 5-year rates of OS and PFS were 38.1% and 23.2%, and 25.9% and 17.9%, respectively.
Kaplan-Meier curves of overall survival (A) and progression-free survival rates (B) and patterns of failures following salvage radiotherapy for locoregionally recurrent esophageal cancer (C).
Distant metastasis was the most common failure pattern following salvage RT-based treatment, occurring in 59 patients (40.1%) (Fig. 1C), and the median distant metastasis-free survival was 12 months. In-field loco-regional progression with or without out-field/distant progression was observed in 26 patients (17.7%).
3. Prognostic factors
In the univariate analysis for survival outcomes, advanced initial T category (T2-4) (p=0.019) and lower BED10 (< 78 Gy, p=0.035) were associated with inferior OS (Table 2). A factor associated with inferior PFS was shorter DFI (< 24 months, p=0.037), and there was a trend of inferior PFS in the initial N2-3 category (p=0.092) (Table 2). In multivariate analyses, there was no significant factor for OS or PFS (Table 2).
Univariate and multivariate analysis for overall survival and progression-free survival
4. Toxicities
Tables 3 and 4 summarizes the treatment-related toxicity profiles, and the most common acute toxicity was esophagitis observed in 68.7% of the patients. The most common late toxicity was esophageal stricture observed in 11 patients (7.5%), and grade 3 or higher late esophageal or tracheal toxicities were observed in seven patients (4.8%). In addition, grade 4 toxicity occurred in four patients: esophageal perforation in one; and fistula in three, respectively.
Overall incidences of treatment-related toxicities
Incidences of grade 3 or higher toxicities in relation to the radiation technique, BED, and concurrent chemotherapy
Overall, 16 patients (10.9%) experienced grade 3 or higher toxicities, which was more frequently observed among the patients who underwent CCRT (19.6% vs. 6.3%, p=0.023). Meanwhile, the incidence of grade 3 or higher toxicity was similar according to BED10 levels (< 78 Gy vs. ≥ 78 Gy; 14.3% vs. 8.8%, p=0.414) (Table 4).
5. Risk stratification
In order to identify the significance of risk factors associated with OS or PFS, comparative analyses were done on the potential risk factors including initial T2-4 category, initial N2-3 category, and DFI < 24 months, respectively. The disparity in the outcomes was most prominent when the patients had two or more of three potential risk factors, who were defined as high-risk group (S2 Table). In a multivariate analysis that incorporated the risk group, the high-risk group was significantly associated with significantly inferior OS (p=0.020) and PFS (p=0.043), respectively (Table 5).
Cox proportional hazard multivariate model for survival outcomes
Discussion
In this study, we retrospectively analyzed the treatment outcomes following salvage RT-based treatment for LRR following upfront surgery in the esophageal cancer patients. Although several previous studies reported the survival outcomes in this setting, the majority was in retrospective study nature, which included mostly dozens of patients. A systemic review including 1,553 patients from 27 retrospective and three prospective studies showed that the 1-, 3-, and 5-year OS rates were 67.9%, 35.9%, and 30.6%, respectively [15]. In the current study, the corresponding OS rates of 147 patients were 66.7%, 31.2%, and 23.2%, respectively, which seemed not very different from those reported in the systematic review. Though objective comparison was not feasible, our data seem to confirm the previous reports.
As the first failure site following RT-based salvage treatment, approximately 40% of the patients developed distant metastasis in our study, with the median time to distant recurrence of 12 months. Meanwhile, one-fourth of the patients showed long-term survival over 5 years (Fig. 1), which may suggest that there exist an innate biological heterogeneity among the patients. To the best of our knowledge, however, there is no established biomarker that could predict the survival outcome or early distant recurrence in primary or recurrent esophageal cancer. Mummundi et al. [15] suggested that the early detection of recurrence through regular imaging examination and advanced imaging studies such as FDG-PET/CT might improve survival because only 6.9% of the patients included in their meta-analysis performed PET-CT at recurrence [15]. Imaging evaluation, however, may have limited capability in predicting early distant recurrence or detecting microscopic disease considering that 40% of our patients developed distant metastasis though FDG-PET/CT was done and showed no evidence of distant metastasis in 84.3% of the patients at the time of recurrence. The current study intended to stratify the risk group based on the initial T category, N category, and DFI, which were known as the risk factors for recurrence, and we found that the high-risk group was associated with significantly inferior OS and PFS. These clinical factors could serve as prognosticators and/or a guide for more-aggressive treatment approach.
There were a few studies that reported the benefit of concurrent chemotherapy in addition to salvage RT for recurrent esophageal cancer [9,13,15]. The NCCN clinical practice guidelines preferred CCRT in LRR setting following esophagectomy if the patients did not undergo CCRT previously. In our study, CCRT showed a trend to improve PFS (p=0.090), but failed to improve OS (p=0.336) (Table 2), with the price of worse toxicity profiles (Table 3). Yamashita et al.’s Japanese multi-institutional study [9] showed a significant OS benefit by CCRT on multivariate analysis, where the 5-year OS rates were 0% in RT alone arm and 31% in the CCRT arm, respectively. In contrast, the current study showed that the 5-year OS rates were 22% in RT alone arm and 25% in the CCRT arm, respectively. These discrepancy might reflect that there could have been the selection bias in Yamashita et al.’s [9] and our studies, both of which were retrospective ones. In this study, the patients who underwent CCRT more frequently had multiple lesions and were younger than those who did RT alone (Table 1). This implies that the benefit of concurrent chemotherapy needs further evaluation through a larger-scale prospective study.
The optimal RT dose in salvage treatment for LRR of esophageal cancer has been still controversial. Although there were several studies that supported improved survival outcomes with increased radiation dose, there were some other studies, however, that failed to prove this high-dose benefit [14,17-19]. A recent meta-analysis reported that higher RT dose (> 50 Gy) was a significant factor for improved survival in the recurrent esophageal cancer patients [15]. In the current study, higher RT doses (BED10 ≥ 78 Gy) were associated with improved OS in univariate analysis, whose significance, however, disappeared in multivariate analysis. Meanwhile, high-dose RT (BED10 ≥ 78 Gy) was not associated with more frequent occurrence of severe toxicities (Table 2). In the current study, we used the modern RT techniques such as 3D-CRT and IMRT, and 97.3% of the patients received BED10 ≥ 70.2 Gy (EQD2 > 58 Gy), which was high enough when compared to the suggested dose of 50 Gy in the previous meta-analysis. Our results showed acceptable toxicity profiles (less than 5% of grade 3 or higher late esophageal or tracheal toxicities), when considering high dose was delivered to the re-located stomach in some of our patients. Our findings suggest that hypofractionated RT with 54-60 Gy administered at 3 Gy per fraction might not be too dangerous. However, there are concerns about whether the re-located stomach can endure these high dose. Liu et al. [20] reported that the intrathoracic stomach could tolerate 60 Gy or more in the patients who underwent esophagectomy. The tolerability of the re-located stomach might differ, for example, due to the absence of gastric acid secretion. However, given the retrospective nature of this study, a high level of caution is needed when administering high doses to the re-located stomach. Further research is desired to determine the optimal salvage radiation dose for managing recurrent esophageal cancer and to understand how surgical re-location affects the stomach’s radiation tolerance.
This study has a few limitations. Due to the retrospective nature, there might have been a selection bias which could not be easily overcome though we performed multivariate analysis. In addition, there might have been stage migration issue, because of the long study period as well. Due to the long study period, half of the patients in this study, who were managed during the early study period, did not receive the initial treatment pertaining to the current standard of care. To assess whether the outcomes have varied over the prolonged study period, we analyzed whether there were differences in survival outcomes according to the year of treatment or in rates of toxicities according to the RT techniques, as detailed in Tables 2 and 3. Our analysis found no significant difference on these factors. Nevertheless, our study has the merit of a large number of patients who underwent salvage RT-based treatment at a single institution. Our results suggest that high-dose radiation in salvage RT or CCRT for LRR of esophageal cancer could be a feasible and safe option. Further studies, however, are warranted to further clarify the patients’ subgroup who may benefit more. In conclusion, salvage RT alone as well as CCRT seem effective in the patients with LRR of esophageal cancer following upfront surgery.
Electronic Supplementary Material
Supplementary materials are available at Cancer Research and Treatment website (https://www.e-crt.org).
Notes
Ethical Statement
This study was approved by the Institutional Review Board of Samsung Medical Center (SMC-2022-11-131) and conducted in accordance with the principles of the Declaration of Helsinki. Informed consent was waived because of the retrospective nature of the study and the analysis used anonymous data.
Author Contributions
Conceived and designed the analysis: Oh D, Ahn YC, Sun JM, Kim HK, Shim YM.
Collected the data: Cho WK, Noh JM, Oh D, Ahn YC, Sun JM, Kim HK, Shim YM.
Contributed data or analysis tools: Cho WK, Noh JM, Oh D, Ahn YC.
Performed the analysis: Cho WK, Noh JM.
Wrote the paper: Cho WK, Noh JM.
Conflict of Interest
Yong Chan Ahn, the editor-in-chief of the Cancer Research and Treatment, was not involved in the editorial evaluation or decision to publish this article. All remaining authors have declared no conflicts of interest.