Effectiveness of two high-dose-rate intraluminal brachytherapy schedules for symptom palliation in carcinoma esophagus: A tertiary care center experience
Aim: The aim was to analyze different radiation schedules with high-dose-rate (HDR) brachytherapy in patients with unresectable carcinoma esophagus in terms of dysphagia-free survival (DyFS), local control (LC), disease-free survival (DFS), and complications.
Keywords: Carcinoma esophagus, Disease-free survival, Dysphagia-free survival, High dose rate, Local control
Esophageal cancer is the third most common cancer in India constituting 6-7% of all newly diagnosed cancers.  Though currently, surgical excision is the primary modality of treatment, but it is not always applicable because 80% patients present in an advanced stage or comorbid conditions or because of patient's unwillingness or lack of surgical expertise. Radical external radiation alone results in a dismal 5-year survival rate of 6%, with more than 80% failing locally only. Still, the 3-year survival hover around 17-20% only. ,,, Because of the impressive 3-year survival results (27-30%), concomitant chemoradiation is now considered as the standard approach for the treatment of unresectable carcinoma esophagus. Despite these improved results, the locoregional control of the disease is poor, and about 48% patients fail locally. 
The key to further improvement in the treatment results therefore appears to lie in increasing the biological response with the optimal use of combined modality treatment, namely, radiation only or chemoradiation followed by intraluminal brachytherapy (ILBT). ILBT, due to its rapid dose fall-off and ability to deliver a high dose to the intraluminal disease, is used for local boost. Because of the shorter treatment time, better patient compliance, and ease of administration, high-dose-rate (HDR) brachytherapy is preferred. But the dose fractionation and dose per fraction in HDR brachytherapy of carcinoma esophagus has been a matter of concern because of the high incidence of adverse events with certain dose schedules. , This retrospective study has been done with a premise to analyze the different radiation schedules with ILBT in patients with unresectable carcinoma esophagus in terms of dysphagia-free survival (DyFS), local control (LC), disease-free survival (DFS), and complications.
This is a retrospective study of patients of esophageal carcinoma treated between January 2005 and June 2010. Patients were considered eligible for the study if they had squamous cell carcinomas of thoracic esophagus, without distant metastasis, a Karnofsky performance score (KPS) more than 70, a tumor length less than 10 cm, and undergone external beam radiation therapy with or without concurrent chemoradiation, followed by ILBT [Figure 1]. Out of the total 1580 patients of esophageal carcinomas registered in the institution, 86 patients (5.4%) were selected who could satisfy the above-mentioned criteria.
Initially at our institution, patients were treated with external beam radiation therapy alone (schedule A), with no chemotherapy, followed by 2-3 weeks of HDR brachytherapy. The total biological equivalent dose (BED) of this schedule was 98.22 Gy 3 , where Gy 3 stands for the dose related to late toxicities to normal tissues with an α/β ratio of 3. Later on to improve the results of LC, the EBRT dose was increased from 35 Gy/15# to 44 Gy/22#, and concurrent weekly chemotherapy was added. After concurrent chemoradiation, patients were treated with two different HDR brachytherapy schedules.
Both these schedules, schedule B (6 Gy×2) and schedule C (4.68 Gy×3) were designed to be equivalent for late effects using the linear quadratic model and the BED for both these schedules was 109.33 Gy 3 .
Patients were retrospectively analyzed. [Figure 2] shows the distribution of the patients across the three radiation schedules.
All patients included in the analysis underwent initial evaluation with a barium swallow and endoscopy with biopsy to assess the mucosal extent of the disease and a CECT scan to assess the extraesophageal and nodal spread. The patient and tumor characteristics have been shown in [Table 1].
Treatment planning and delivery
All patients were planned on a simulator CT and following a barium swallow, fields were set so as to cover the tumor adequately with a margin of 5 cm proximally and distally and 2-3 cm laterally. Parallel opposing anteroposterior fields were used. All patients were treated on Theratron 780 C or 6 MV LINAC.
For schedules B and C, the concurrent chemotherapy consisted of weekly cycles of cisplatin 30 mg/m 2 in infusion with adequate hydration and a 5-fluorouracil 325 mg/m 2 IV bolus. Weekly hemogram and renal function tests were checked prior to chemotherapy. On the completion of chemoradiation, 2-3 weeks were allowed for the resolution of acute radiation reactions.
ILBT was performed on an outpatient basis. Patients were given injection pentazocine 30 mg/m 2 IV and injection hyoscine butyl bromide 20-40 mg/m 2 IV as premedication. Endoscopy was done by the gastroenterologists with a flexible fiberoptic endoscope (Olympus GIF 160). The parameters evaluated included the extent of the tumor, presence of ulcer, and stricture. A Pagliero Rowland type Selectron bougie, 100 cm long and with 1 cm external diameter, was then threaded over a guidewire and positioned so that the lower end was 2 cm beyond the lower end of the initial lesion.
Orthogonal x-rays were taken to map the shape of the inserted catheter. Treatment planning was done using the PLATO treatment planning system (Nucletron). The dose was prescribed at 1 cm from the central axis of the source as per American Brachytherapy Society Consensus Guidelines.  Patients were treated using a remote afterloading machine (Microselectron) using an Ir 192 source with a maximum activity of 10 Ci.
Follow-up and evaluation
Patients were clinically evaluated for dysphagia at 1, 2, 3, 6, and 12 months. A barium swallow was done in the first and third month. Endoscopy was done at 3 months for disease assessment and biopsy was done when indicated. CECT of the chest and abdomen was obtained if any clinical suspicion of either local recurrence or metastasis was there.
Toxicities were graded as per the RTOG/EORTC criteria, and the toxicities analyzed are shown in [Table 2]. The median follow-up period was 12.1 months.
Definition of the response
Response to treatment was evaluated at 3 months. A complete response was defined as a complete disappearance of the endoscopically visible disease and the absence of malignant cells on biopsy, and a partial response as a reduction in tumor mass by more than 50% without new areas of tumor development. A minimal response was defined as a reduction of less than 50% in the size of the tumor while progressive disease as an increase in the size by more than 25% or appearance of new lesions.
Statistical analysis was performed using the Statistical Package for Social Sciences (SPSS) software, version 16.0. Survival analysis and actuarial probabilities were calculated with the Kaplan-Meier test. All P-values were two sided, and a P-value of 0.05 was taken as significant.
The treatment was well tolerated and all patients completed the course of treatment prescribed. None of the patients had required a treatment gap during the course of external radiation. There were no significant differences in the incidence of acute toxicities across the three radiation schedules. In treatment toxicities as shown in [Table 2], no grade III toxicities were noticed.
Four patients, two each in schedules A and B, had required gaps between the ILBT sessions due to the presence of deep ulcers. This gap was required after the first session in both patients. In both these patients, ILBT was done after a period of 2 weeks.
[Figure 3] shows the prevalence of dysphagia in the three radiation schedules across the follow-up period. The maximum incidence of dysphagia was seen at 1 month with all the three radiation schedules, the highest being 95%, seen among those treated with the radiation only schedule, i.e., schedule A. The incidence of dysphagia at 1 month was 90% and 80% for schedules B and C, respectively. The absence of dysphagia was noted in 40%, 50%, and 55% patients in the three schedules, respectively, analyzed for outcomes post-treatment in the sixth month. However, the incidence of patients with dysphagia increased again at 12 months, and the increase was maximally seen among schedule A (77%), followed by schedule B (72%). Only 55% patients treated with schedule C had dysphagia at 12 months.
All patients had a partial response locally after external radiation. Following ILBT, the radiological response was assessed at 1 month with a barium swallow. [Figure 4] shows that there was no significant difference in the partial response rates in the three radiation schedules. The complete response rate was 39% in schedule A, 40% in schedule B and 44% in schedule C (P=0.6).
[Figure 5] and [Table 3] show that the median DyFS is the highest, 12.1 months, in schedule C, while for schedules A and B (higher dose per fraction) it is low, i.e., 5.1 and 7.0 months, respectively. The 2-year actuarial DyFS is again insignificantly higher for schedule C as compared to schedules A and B (P=0.168). It increased by 13-18% with chemoradiation schedules compared to radiation alone schedules.
As shown in [Figure 6] and [Table 4], the 2-year LC rates were the worst with radiation alone schedules, i.e., 67.6% after a median follow-up of 12.1 months, while with chemoradiation schedules, it was approaching 88%. However, the difference seen is not significant (P=0.48).
[Figure 7] and [Table 5] show that the 2-year DFS was again insignificantly higher for chemoradiation schedules, while for the radiation only schedule, i.e., schedule A, it is low (only 46%).
[Figure 8] shows that the stricture formation was the commonest complication observed within all the three radiation schedules, but it was the highest, seen in 45% (20/51) patients treated with radiation only schedule A, and 21% (5/21) in schedule B, i.e., higher dose per fraction of brachytherapy. It was the lowest in schedule C, 21% (2/14). Ulceration was seen in two patients each in schedules A and B. Tracheoesophageal fistula developed in total five patients, two each in both chemoradiation schedules and one in the radiation only schedule. Hemorrhage was seen in one patient in schedule B. There was no complication in 71% patients (10/14) treated with schedule C.
[Figure 9] shows the pattern of failure at the time of the last follow-up. Local recurrence was seen in 30 patients out of 51 in schedule A, while it was seen in 4 of 21 and 2 of 14 patients in schedules B and C, respectively. A total of 13 patients died of brain metastasis among 86 patients, and 11 among these were treated with schedule A. Seven patients developed lung metastasis. A total of 7 patients among 81 developed metastasis at sites other than brain and lung.
External radiation alone without chemotherapy followed by ILBT in esophagus has been considered to be associated with lower LCs and DFS in various studies. To improve upon the results, few studies have been done on ILBT following chemoradiation, and have reported differing toxicities and clinical benefits. ,, Based on these studies, the American Brachytherapy Society had given consensus guidelines on the use of ILBT.  They had suggested a dose/fractionation of 10 Gy by HDR brachytherapy given in two sessions of 5 Gy each 1 week apart. Clinical data from patients treated at our institute using two sessions of 6 Gy each had shown good LC, with acceptable toxicities. Hence, the radiation only schedule (with higher dose per fraction of external radiation) has been compared with two biologically equivalent chemoradiation schedules (one with a lower dose per fraction and the other with a higher dose per fraction of brachytherapy boost) to analyze the three schedules in terms of DyFS, LC, DFS, and toxicities.
Thirty-five patients treated with chemoradiation completed the scheduled course with at least three cycles of chemotherapy followed by ILBT. Though grade I and II hematological toxicity, renal toxicity, and vomiting were seen, no grade III toxicity was noted. Western studies however quote a poorer compliance to treatment of 64-70% only, , with grade III and IV toxicities noted in 30-58% patients, probably due to the higher doses of external radiation and chemotherapy used in these studies. These results show the feasibility of the treatment schedule.
Across the three schedules, dysphagia scores were initially high in all the schedules, which steadily decreased and plateaued off. However, grade II and III dysphagia appeared to increase in the sixth month in schedules A and B. Although not statistically significant because of the small patient number, this might be reflective of the slightly higher rate of stricture formation in both schedules A and B due to the higher dose per fraction of external radiation and higher dose per fraction of brachytherapy boost, respectively. Schedule C shows a trend toward slightly better symptom control, which is supported by the evidence of a significant increase in the number of patients without dysphagia in the post-treatment period at 1 month and also the highest median DyFS of 12.1 months.
Schedules B and C showed a trend toward better LC, 87.5% and 88.9% versus 67.6% in the radiation group, respectively (P value, not significant), but major complications like ulceration and tracheoesophageal fistulas were seen more commonly in schedule B using higher dose per fraction of brachytherapy boost. These results are similar to other published studies, which used higher doses of external radiation and chemotherapy.  In the RTOG 92-07 trial, in which patients were given 50 Gy external radiation, followed 2 weeks later by ILBT - 2 sessions of 5 Gy each and chemotherapy on weeks 1, 5, 8, and 11, the LC rate was 73%.  However, there was a high incidence of severe complications (58%), including 12% treatment-related esophageal fistula, based on which, an extreme caution in employing ILBT boost following chemoradiation was urged. A study by Calais et al., in which 60 Gy conventional external radiation was given concurrent with cisplatin, 5-flourouracil, and mitomycin followed by two sessions of ILBT 5 Gy each, reported LC rates of 74% at 1 year. This later came down to 57% at 3 years. This indicates that chemoradiation schedules with brachytherapy boost are feasible but require a careful selection of patients with a good KPS, along with a need of further studies to standardize HDR ILBT schedules with chemoradiation protocols.
In patients with unresectable carcinoma esophagus, concurrent chemoradiation followed by an ILBT boost is feasible with an acceptable toxicity profile and good LC rates. However, an increased risk of complications is seen with the higher dose per fraction of brachytherapy (6 Gy×2#). Radiation only schedules with ILBT are associated with lower LC and DFS. So wherever suitable, we should try to add chemotherapy concurrent with radiation and boost it with ILBT. Still there is a need for the standardization of HDR ILBT schedules with chemoradiation protocols.
[Figure 1], [Figure 2], [Figure 3], [Figure 4], [Figure 5], [Figure 6], [Figure 7], [Figure 8], [Figure 9]
[Table 1], [Table 2], [Table 3], [Table 4], [Table 5]