Increased dose of re-irradiation therapy or combined chemotherapy to improve survival of local recurrent esophageal squamous cell carcinoma after radiotherapy

Introduction: This study aimed to explore the optimal treatment for locally recurrent esophageal squamous cell carcinoma (LRESCC) by analyzing the results of different doses of re-irradiation therapy after radical (chemo) radiotherapy. Methods: We retrospectively analyzed 125 patients with LRESCC after initial radiotherapy. According to radiotherapy dose, 58 patients were assigned to the low-dose (LD) group (50–54 Gy) and 67 to the high-dose (HD) group (55–60 Gy). We recorded the response rate (complete + partial response), 1-, 2- and 3-year survival rates, and toxicity in the two groups. The effects of different radiotherapy doses and combination chemotherapy on the survival of LRESCC were analyzed. Results: After re-irradiation, the 1-, 2- and 3-year survival rates were 48.3%, 24.1% and 10.3% in the LD group and 61.2%, 34.3% and 19.4% in the HD group (P<0.05), respectively. The median survival time of patients receiving radiotherapy alone was 9 months in the LD group and 15 months in the HD group (P<0.05). The survival rate of patients treated with chemoradiotherapy was higher than that of patients treated with radiotherapy alone in the LD group, but in the HD group, chemoradiotherapy showed no advantage compared with radiotherapy alone. The most common toxicity was esophagitis, and the incidence was higher in the HD group than in the LD group (68.7% vs 58.6%, P (cid:0) 0.05). Conclusion: Increasing the dose of re-irradiation therapy or chemotherapy can improve the ecacy of re-irradiation therapy for patients with LRESCC after radiotherapy.


Introduction Esophageal squamous cell carcinoma (ESCC) is the most common esophageal malignant tumor in
China. It is moderately sensitive to radiation. Comprehensive treatment based on radiotherapy or surgery is its main treatment. Unfortunately, the local recurrence rate of esophageal cancer is as 40%-60% after radiotherapy [1] . To date, there is still no consensus on the retreatment of local recurrent ESCC (LRESCC) after (chemo)radiotherapy. Studies have shown that salvage esophagectomy is effective in treating locally recurrent esophageal cancer [2,3]. Chen et al. reported that re-irradiation can achieve similar e cacy to salvage surgery [4] . In addition, brosis of the tissue surrounding the tumor bed after radiotherapy increases the di culty of surgery and perioperative risk. Therefore, even carefully selected patients should be cautiously considered for surgery [5,6].
There are still many problems in the re-irradiation of LRESCC. Radiotherapy dose is an important factor affecting e cacy. There is no consensus on the implementation of re-irradiation dose. Chemotherapy is often used as a systemic treatment, but in the case of repeat radiotherapy for LRESCC, the role of combined chemotherapy is not clear. Therefore, the purpose of present study was to evaluate the e cacy and toxicity of different radiotherapy doses and combined chemotherapy with re-irradiation therapy for LRESCC.

Patients and clinical features
We retrospectively selected 207 patients with LRESCC who underwent re-irradiation therapy in the Chengdu Fifth People's Hospital between January 2012 and December 2016. The inclusion criteria were: 1) non-recurrence interval > 6 months after the end of initial radiotherapy; 2) pathological examination, imaging or gastroscopy con rmed the recurrence of primary esophageal cancer with or without local lymph node metastasis, or only local lymph node metastasis (supraclavicular fossa, mediastinal, esophageal, or para-aortic lymph nodes); 3) re-irradiation therapy dose 50-60 Gy; 4) Eastern Cooperative Oncology Group (ECOG) score 0-2, and good function of heart, liver, kidneys, lungs, and bone marrow hematopoiesis. The exclusion criteria were: 1) distant organ metastasis, tumor in other organs, or incomplete information or hyper-fractionated radiotherapy; 2) a history of surgical resection of esophageal cancer. Finally, 125 patients were included, 84 men and 41 women with a median age of 68 years (range 50-89 years). This study was approved by the Ethics Committee of Chengdu Fifth People's Hospital.

Treatment and follow-up
In the initial treatment, two-dimensional radiotherapy, three-dimensional conformal radiotherapy (3D-CRT) or intensity modulated radiotherapy (IMRT) was used for radical radiotherapy for the 125 patients, with a median dose of 60 Gy (50-66 Gy). All patients did not undergo surgery. Most patients received chemotherapy regimens containing cisplatin, paclitaxel or uorouracil. After initial treatment, gastroscopy, esophageal barium meal radiography, or enhanced computed tomography (CT) was used to evaluate the e cacy, and complete response or partial response was achieved in all patients.
All patients were re-irradiation with 3D-CRT or IMRT with 1.8-2.0 Gy per fraction, 5 days/week. Gross tumor volume (GTV) was observed on esophageal barium meal examination and CT images. GTV of the metastatic lymph nodes (GTVnd) included enlarged lymph nodes with a short diameter > 10 mm, or positron emission tomography-CT examination suggested metastatic lymph nodes. Planning target volume (PTV) was formed by extending 0.5-1.0 cm radially from GTV, 1.0-2.0 cm above and below GTV, or 0.5 cm outside GTVnd in all directions. The lymphatic drainage area was not irradiated prophylactically. Dose limitation of normal tissue or organ was: bilateral lungs V20 < 25%, spinal cord Dmax < 20 Gy, and mean radiation dose to the heart < 30 Gy.
The overall survival (OS) was de ned as time from the beginning of retreatment until patient death or last follow-up. E cacy was evaluated at the end of retreatment and 1 month after the end of retreatment. Evaluation methods included endoscopy, tumor markers, esophageal barium meal, and enhanced CT. The response rate (RR) was de ned as the percentage of patients with complete response and partial response in the total number of cases. Therapeutic evaluation criteria refers to response evaluation criteria in solid tumors (RECIST 1.1). Toxicity was graded by the Common Terminology Criteria for Adverse Events (CTCAE v4.0). According to the re-irradiation dose, 58 patients were assigned to the low-dose (LD) group (50-54 Gy, median dose 50 Gy) and 67 to the high-dose (HD) group (55-60 Gy, median dose 60 Gy).

Statistical analysis
Statistical analysis of the database was performed using SPSS version 23.0 (IBM Corporation, Armonk, NY, USA). The continuous and categorical variables in the two treatment groups were compared by t-test and chi-square test, respectively. OS was calculated from the rst day of retreatment to the patient's death or last follow-up. The survival rates were calculated by Kaplan-Meier method, and the difference in survival curves was compared by log-rank method. Cox's proportional hazards regression model was used to determine the effect of univariate and multiple factors on survival. Non-recurrence interval (NRI) was de ned as the time between the beginning of initial treatment and diagnosis of recurrence. Twosided P < 0.05 was considered to indicate statistical signi cance.

Results
The clinical baseline characteristics of the two groups of patients are shown in Table 1. The median time to NRI in all patients was 21 months (range 8-201 months). Sixty-eight (54.4%) patients relapsed within 2 years after initial treatment. The last follow-up time was December 2019 with a median follow-up of 19 months (range 4-65 months). At the end of the follow-up, two patients survived in the LD group, seven survived in the HD group; six of the nine patients had a second recurrence, and three did not have a second recurrence or metastasis. There were 67 cases of primary recurrence (PR), 33 cases of primary recurrence accompanied by local lymph node recurrence (PR + LNR) and 25 cases of local lymph node recurrence (LNR). Sixty-one of the 125 patients received concurrent or 1-4 sequential cycles of chemotherapy, which included either single or double-drug regimens containing paclitaxel, platinum or uorouracil.  (Fig. 2). After retreatment, the RR was 87.9% in the HD group (58/66), which was higher than 76.8% in the LD group (43/56), the difference was not signi cant (P = 0.065). The 3-year locoregional control rates in the LD group and that in the HD group were 3.4% and 14.9%, respectively (Fig. 1). There was a signi cant difference between local control rates in the two groups (log-rank test, p = 0.039).
In strati ed analysis, median survival time was 15 months (95% CI: 7.365-22.635 months) in 31 patients in the HD group who received radiotherapy alone. This was superior to median survival of 9 months (95% CI: 7.124-10.876 months) in 33 patients in the LD group who received radiotherapy alone (Fig. 3A). The difference between the two groups was signi cant (P = 0.009). However, in the strati ed analysis of patients received chemotherapy in the two groups, patients in the HD group did not show survival bene t ( Fig. 3B).
In subgroup analysis, we found that chemotherapy had a signi cant effect on survival in the LD group. In the LD group, patients with chemoradiotherapy showed superior median survival than patients with radiotherapy alone (14 months, 95% CI: 12.378-15.622 months vs 9 months, 95% CI: 7.124-10.876 months) (Fig. 4A). The difference was signi cant. However, chemotherapy did not continue to increase the survival advantage of patients in the HD group (Fig. 4B).

Cox regression analysis for overall sample
We conducted a multivariate analysis in order to examined whether the dose of re-irradiotherapy was a factor affecting the prognosis of patients. The results of univariate and multivariate analyses for OS were summarized in Table 3. ECOG of 0-1 (P = 0.048), age 70 years (P = 0.028), NRI 24 months (P = 0.001) and re-irradiation dose of 55-60 Gy (P = 0.013) were associated with favorable OS by Cox analysis. Esophagitis was the most common toxicity in the two groups. The incidence of esophagitis was 58.6% in the LD group and 68.7% in the HD group. The difference was not signi cant (P = 0.868). Other toxicity included hematological toxicity, the difference was signi cant (p = 0.004)and gastrointestinal reaction, the difference was not signi cant (p = 0.732). Rare adverse effects included radiation-induced pneumonitis, radiation tracheitis and skin reaction ( Table 2). Gastrointestinal reactions include nausea, vomiting and loss of appetite. Hematological toxicity was mainly reduced white blood cells and neutrophils, and anemia. After symptomatic treatment, all patients completed radiotherapy. Eleven patients developed severe treatment-related toxicity, including esophageal perforation/ stula and bleeding. The NRI of most of the 11 patients (9 / 11) was ≤ 24 months. After six months of retreatment, two patients in each group had severe esophageal stenosis and underwent esophageal dilatation.

Discussion
External beam radiotherapy plays an important role in relieving the symptoms of esophageal cancer and maintaining quality of life of patients, but its effects are limited. Most patients have recurrence or metastasis within 2 years after radiotherapy [4,7,8]. Without active intervention, the survival and prognosis of patients with disease progression are often poor, and most die within 1 year [9]. Although many researchers are actively exploring the best salvage treatment for LRESCC, there is still no consensus. Many studies have con rmed the positive role of chemotherapy in the initial treatment of ESCC [10], but the role of chemotherapy in the re-irradiation therapy of ESCC is not clear. The clinical e cacy of recurrent ESCC salvage radiotherapy is signi cantly affected by the radiation dose [11]. Up to date, the optimal dose of salvage radiotherapy is still unclear.
Recently, a study of the effect of dose-escalated radiotherapy on ESCC showed that, in patients who achieved complete response, radiotherapy dose was increased to > 59.4 Gy after standard concurrent chemoradiotherapy (50.4 Gy), which signi cantly improved local control, 5-year recurrence-free survival rate and 5-year OS rate [12]. Previously, some studies have indicated that the survival rate of LRESCC can be signi cantly prolonged when the dose of re-irradiation is > 50 Gy [4,13]. Kobayashi et al. pointed out that 60 Gy was an appropriate salvage dose for LRESCC after surgery [14]. Another study from Japan pointed out that patients with LRESCC with a radiation history should be given higher doses of radiotherapy [15]. These studies suggest that increasing the dose of radiotherapy for recurrent esophageal cancer may be bene cial to prolong survival. However, there is no report on whether higher doses of re-irradiation between 50 and 60 Gy would bene t more patients with LRESCC with radiation history.
This study is believed to be the rst to compare the effects of 50 Gy and 60 Gy re-irradiation dose on the survival of patients with LRESCC with radiation history. Our results showed that most patients (68/125, 54.4%) had recurrence within 2 years after initial (chemo)radiotherapy, which was consistent with previous reports [4,8,13,16]. Jingu et al. reported that the 3-year survival rate and median survival time of 33 patients with LRESCC receiving 60 Gy (range, 18-70 Gy) re-radiotherapy were 17.8% and 16 months, respectively [15]. Hong et al. reported that the median survival time of recurrent ESCC patients receiving 50 Gy (range, 21-70 Gy) re-radiotherapy was 10 months [17]. This is close to our survival results. In the present study, the 1-and 3-year survival rates and median survival time were 61.2% and 19.4% and 18 months in the HD group, respectively. The 1-and 3-year survival rate and median survival time were 48.3% and 10.3% and 11 months in the LD group, respectively. Radiotherapy dose is an important factor affecting survival, which was clearly demonstrated in our results. The median survival time of patients receiving radiation alone in the HD group was 15 months, which was signi cantly higher than 9 months in the LD group. Increasing the dose of radiation showed an encouraging outcome. Our results are also better compared with those of other studies [4,13,15]. The reason may be that higher doses of radiation were more bene cial for tumor control. Therefore, we recommend higher doses of radiation for patients with recurrent ESCC after radiotherapy.
Few studies have addressed the role of chemotherapy in re-radiotherapy for recurrent ESCC. Chen et al. reported that the 1-, 2-and 3-year survival rates of 36 LRESCC patients receiving re-irradiation with concurrent chemotherapy (paclitaxel + cisplatin) were 51.7%, 21.4% and 12.2%, respectively [4]. Katano et al. reported that the median survival time of six patients who underwent concurrent chemotherapy with nedaplatin and tegafur with re-radiotherapy was 13.6 months [6]. As of now, the role of chemotherapy in re-radiotherapy for recurrent ESCC is still unclear. In subgroup analysis, our results showed that chemotherapy combined with radiotherapy could increase survival rate. In the LD group, the median survival time for chemoradiotherapy was 14 months, which was consistent with the results reported by Katano et al. [6]. However, in the HD group, chemoradiotherapy did not show better e cacy than radiotherapy alone, and other studies have reported the same [17]. Our results suggest that chemotherapy can increase the survival rate of re-radiotherapy for ESCC, especially when patients are exposed to lower radiation doses. This may be related to the fact that chemotherapy increases the tumor control rate at low rather than high dose of radiotherapy.
In this study, the most common toxicity of the whole cohort was esophagitis. Although the incidence of esophagitis in the HD group was higher than in the LD group (68.7% vs 58.6%), the difference was not signi cant. A previous study showed that the incidence of severe acute esophagitis in patients with thoracic radiotherapy was 15-25% [18]. Grade 3 or higher acute esophagitis was rare in the two groups in this study. Hematological toxicity and gastrointestinal reactions were also common, and only hematologic toxicity was signi cant between the two groups. In this study, the incidence of radiation pneumonitis was lower than in previous studies [19]. Grade 1-2 radiation pneumonitis occurred in six (10.3%) and eight (11.9%) patients in the LD and HD groups, respectively, and the symptoms was controllable. We recorded eleven patients with severe treatment-related toxicity, including bleeding, esophageal perforation, and esophageal stula, and eight of eleven were found in the LD group. This may be related to the NRI of ≤ 24 months in most patients (36/58, 62.1%) in the LD group. Current results indicated that a higher re-irradiation dose was safe.

Conclusion
We recommend that a higher re-irradiation dose (55-60 Gy) for LRESCC patients with radiation history may be more conducive to long-term survival, and toxicity is tolerable. Re-irradiation combined with chemotherapy can improve survival rate of patients, especially those receiving lower doses of reirradiation (50-54 Gy). For those patients with a short recurrence time (NRI ≤ 24 months), we should be vigilant for serious toxicity that may lead to rapid death during re-irradiotherapy.  Figure 1 There was a signi cant difference between locoregional control rate in the LD group and HD group (Kaplan-Meier method)

Figure 3
A Survival curve of patients in the low-dose and high-dose groups after with re-irradiotherapy alone. B Survival curve of patients in low-dose group and high-dose group with chemoradiotherapy. Figure 4