Advertisement

Assessing Outcomes of Patients Treated With Re-Irradiation Utilizing Proton Pencil-Beam Scanning for Primary or Recurrent Malignancies of the Esophagus and Gastroesophageal Junction

Open ArchivePublished:March 04, 2020DOI:https://doi.org/10.1016/j.jtho.2020.01.024

      Abstract

      Introduction

      Re-irradiation (re-RT) for locoregionally recurrent esophageal and gastroesophageal junction (GEJ) cancer and de novo esophageal + GEJ cancer arising in-field after a course of prior radiation poses considerable treatment challenges given the sensitivity of surrounding organs at risk (OARs). Guidelines for treatment of this presentation are not well established. Pencil-beam scanning (PBS) proton therapy has the ability to decrease radiation dose to OARs relative to photon plans. We present the first published series to date of re-RT with PBS for esophageal + GEJ malignancies and hypothesize that re-RT with proton PBS will be feasible and improve the safety profile of re-RT for this cohort of patients.

      Methods

      Consecutive esophageal + GEJ cancers treated with PBS re-RT within a single institution were analyzed. Comparative volumetric-modulated arc therapy photon plans were generated. A total of 17 patients were included for analysis.

      Results

      At a median follow-up of 11.6 months, 1-year local control was 75.3% and overall survival was 68.9%. There were five (27.8%) grade 3 or higher late toxicities. When matched for clinical target volume coverage, proton PBS plans delivered significantly lower doses to the spinal cord, lungs, liver, and heart (all p < 0.05); five volumetric-modulated arc therapy plans would have been undeliverable on the basis of physician-specified OAR constraints.

      Conclusions

      Re-RT for de novo or recurrent malignancies of the esophagus + GEJ, when delivered with PBS proton therapy, yields high rates of local control with acceptable acute and late toxicities in a high-risk population and decreased radiation dose to OARs relative to comparative photon plans. This is the largest series of proton re-RT for esophageal malignancies and the first that exclusively used PBS.

      Keywords

      Introduction

      Cancer of the esophagus and gastroesophageal junction (GEJ) is ranked as the eighteenth most common malignancy in the United States, yet it is the eighth most deadly malignancy, with 15,850 estimated deaths in 2018.
      Surveillance, Epidemiology, and End Results (SEER) Program. 1973–2015, National Cancer Institute, DCCPS, Surveillance Research Program, released Cancer Stat Facts:2017.
      Despite recent improvements in treatments, 5-year overall survival rates still only approach 20%.
      Surveillance, Epidemiology, and End Results (SEER) Program. 1973–2015, National Cancer Institute, DCCPS, Surveillance Research Program, released Cancer Stat Facts:2017.
      After initial management, a subset of these patients will develop locoregional recurrences, including intraluminal or anastomotic plus or minus regional nodal disease, and subsequent treatment-related decision-making becomes complex, particularly in the setting of a previous esophagectomy.
      For these locoregional recurrences, systemic therapy alone is not a curative treatment, and surgical resection is often high-risk, if not anatomically impossible in the instances of esophageal anastomotic recurrences. A second cohort of patients exists who develop de novo esophageal + GEJ cancers as a second malignancy in the setting of a previously irradiated thoracic or head and neck malignancy. Although these malignancies may have a biological profile similar to other de novo esophageal malignancies, they maintain the substantial challenges related to high- dose radiation to the esophagus and nearby organs. Delineating a definitive treatment course for these second malignancies becomes similarly difficult.
      Delivering irradiation (RT) to treat esophageal + GEJ malignancies in a thorax or abdomen that has already seen radiation presents a unique challenge, as the amount of dose that can be delivered is markedly limited by the tolerance of nearby organs at risk (OARs). A paucity of data exists to guide re-RT dosing for either de novo or recurrent esophageal + GEJ malignancies and associated techniques, with treatment being nonstandardized. Published data are either limited by small sample size
      • Fernandes A.
      • Berman A.T.
      • Mick R.
      • et al.
      A prospective study of proton beam reirradiation for esophageal cancer.
      • Katano A.
      • Yamashita H.
      • Nakagawa K.
      Re-irradiation of locoregional esophageal cancer recurrence following definitive chemoradiotherapy: a report of 6 cases.
      • Patel S.A.
      • Edgington S.K.
      • Adams J.
      • Morse C.
      • Ryan D.P.
      • Hong T.S.
      Novel use of proton beam therapy for neoadjuvant treatment of radiation-associated squamous cell carcinoma of the esophagus.
      or focused primarily on photon radiotherapy,
      • Jingu K.
      • Niibe Y.
      • Yamashita H.
      • et al.
      Re-irradiation for oligo-recurrence from esophageal cancer with radiotherapy history: a multi-institutional study.
      often describing a high rate of severe (≥grade 3 [G3]) late treatment-related toxicities.
      • Fernandes A.
      • Berman A.T.
      • Mick R.
      • et al.
      A prospective study of proton beam reirradiation for esophageal cancer.
      ,
      • Kim Y.S.
      • Lee C.G.
      • Kim K.H.
      • et al.
      Re-irradiation of recurrent esophageal cancer after primary definitive radiotherapy.
      In the modern era, with more contemporary treatment techniques, studies have reported a survival benefit when recurrent esophageal cancers are treated with radiation.
      • Hong L.
      • Huang Y.
      • Zhuang Q.
      • et al.
      Survival benefit of re-irradiation in esophageal cancer patients with locoregional recurrence: a propensity score-matched analysis.
      Some emerging studies suggest improved dosimetric and safety profiles when re-RT is given with proton radiation.
      • Fernandes A.
      • Berman A.T.
      • Mick R.
      • et al.
      A prospective study of proton beam reirradiation for esophageal cancer.
      ,
      • Patel S.A.
      • Edgington S.K.
      • Adams J.
      • Morse C.
      • Ryan D.P.
      • Hong T.S.
      Novel use of proton beam therapy for neoadjuvant treatment of radiation-associated squamous cell carcinoma of the esophagus.
      Proton radiation, given its unique physical properties and the nature of its characteristic Bragg peak, more effectively spares intrathoracic and upper abdominal OARs from integral radiation when treating thoracic and upper abdominal targets, which is particularly critical in the setting of re-RT. Most of the published data using proton therapy for re-RT of the esophagus or GEJ has been performed with passively scattered protons.
      • Fernandes A.
      • Berman A.T.
      • Mick R.
      • et al.
      A prospective study of proton beam reirradiation for esophageal cancer.
      ,
      • Berman A.T.
      • Fernandes A.T.
      • Both S.
      • et al.
      Prospective trial of proton reirradiation of locally recurrent esophageal cancer.
      Our institution utilizes pencil-beam scanning (PBS) proton therapy for all patients who have not been previously reported for esophageal or GEJ re-RT. Given the enhanced conformality of proton PBS dose clouds, we hypothesized that re-RT delivered with proton PBS, for locoregionally recurrent esophageal or GEJ cancer and de novo esophageal or GEJ cancer arising in-field after a previous course of radiation would be feasible and well tolerated and would decrease radiation dose to nearby OARs compared with volumetric-modulated arc therapy (VMAT) photon RT.

      Materials and Methods

      After institutional review board approval, a Health Insurance Portability and Accountability Act-compliant database was constructed of consecutive patients diagnosed with de novo or recurrent esophageal or GEJ cancers treated with PBS in a field of previous radiation at a single institution between December 2016 and August 2019. Male and female patients older than or equal to 18 years of age were included if they had a locoregionally recurrent esophageal or GEJ cancer arising in their previous treatment field or a de novo esophageal or GEJ cancer arising within the RT field of a previous malignancy, whether or not this was felt to be a secondary malignancy. Plans were considered re-RT if there was significant overlap of the 50% or higher isodose lines between RT courses. In all cases, the initial course of RT was delivered with photons. Patients were included for analysis if both initial and re-RT courses were delivered with definitive intent. Patient, tumor, and treatment characteristics and toxicities were collected.

      Toxicity Reporting

      Acute and late toxicities were retrospectively analyzed. Toxicities were graded according to the Common Terminology Criteria for Adverse Events v.5.0. Acute toxicities during treatment were graded prospectively on a weekly basis by each patient’s treating physician and retrospectively reviewed; late toxicities were obtained though detailed retrospective electronic medical record review.

      Comparative Plan Generation

      All primary proton PBS plans were generated in the Eclipse treatment planning system, (Varian Medical Systems, CA), on the basis of our institutional standard, and dose calculated with a proton convolution superposition algorithm. The proton plans were subsequently exported into the RayStation treatment planning system (RaySearch Laboratories, Sweden) and anonymized for research purposes; within the RayStation, comparative photon plans were optimized and dose computed using a collapsed cone convolution superposition algorithm and matched to physician-determined prescription clinical target volume (CTV) coverage to ensure that both plans would have delivered adequate target dose. After plan generation, doses to critical OARs, including the spinal cord, total lungs, liver (for distal tumors), and heart, were calculated, and compared with doses which had been delivered with PBS plans. Cumulative OAR dose constraints were determined on an individual basis with physician discretion, taking into account doses received by previous courses of radiation. Comparisons included the ability to successfully meet physician-specified OAR constraints, homogeneity and conformity of target coverage, and differences in objective RT doses received by OARs as part of the re-RT plans.

      Statistical Analysis

      Overall survival, local control (LC), and distant control (DC) were estimated using the Kaplan-Meier method and calculated from the start of re-RT. Log-rank test was used to compare oncologic outcomes between de novo and recurrent lesions on the basis of survival curves generated using the Kaplan-Meier method. The Mann-Whitney U test was used to compare treatment characteristics between de novo and recurrent lesions. Wilcoxon signed rank test was used to compare the doses that would have been delivered to contoured OARs among the plans without compromising target coverage. Statistical significance was defined as p less than 0.05, and analysis was performed using IBM SPSS V24 software.

      Results

      Overall, 17 patients and 18 treatment courses were included for analysis; one patient underwent a total of three courses of radiation for two locoregional recurrences. Patients were predominantly men and all were white, and median age at the time of diagnosis requiring re-RT was 71 years (51–85 y) (Table 1). One patient underwent six fractions of treatment with photons owing to delays with insurance approval; all other patients were treated exclusively with proton PBS.
      Table 1Patient and Tumor Characteristics
      CharacteristicsOverall n = 18De Novo (n = 8, 44.4%)Recurrent (n = 10, 55.6%)
      Gender
      Indicates instance in which overall n = 17 owing to one patient receiving two discreet courses of treatment. When n = 17, de novo lesions n = 8 (47.0%), recurrent lesions n = 9 (53.0%).
       Male16/17 (94.1%)8/8 (100.0%)8/9 88.9%)
       Female1/17 (5.9%)0/8 (0.0%)1/9 (11.1%)
      Race
      Indicates instance in which overall n = 17 owing to one patient receiving two discreet courses of treatment. When n = 17, de novo lesions n = 8 (47.0%), recurrent lesions n = 9 (53.0%).
       White17/17 (100%)8/8 (100%)10/10 (100%)
      Age at treatment start (y)
       Median (range)71 (51–85)72 (59–85)64 (51–78)
      Vital status
      Indicates instance in which overall n = 17 owing to one patient receiving two discreet courses of treatment. When n = 17, de novo lesions n = 8 (47.0%), recurrent lesions n = 9 (53.0%).
       Alive10/17 (58.8%)4/8 (50.0%)6/9 (66.7%)
       Dead7/17 (41.2%)4/8 (50.0%)3/9 (33.3%)
      Lesion status
       De novo8/18 (44.4%)
       Recurrent10/18 (55.6%)
      Previous primary site (if de novo lesion)
       H&N3/8 (37.5%)
       Lung2/8 (25.0%)
       Lymphoma2/8 (25.0%)
       Thyroid1/8 (12.5%)
      Recurrent sites
       Anastomotic6/10 (60.0%)
       Nodal recurrence2/10 (20.0%)
       Anastomotic + nodal2/10 (20.0%)
      Histology
       Adenocarcinoma9/18 (50.0%)0/8 (0.0%)9/10 (90.0%)
       Squamous cell carcinoma9/18 (50.0%)8/8 (100.0%)1/10 (10.0%)
      Vital status
      Indicates instance in which overall n = 17 owing to one patient receiving two discreet courses of treatment. When n = 17, de novo lesions n = 8 (47.0%), recurrent lesions n = 9 (53.0%).
       Alive10/17 (58.8%)4/8 (50.0%)6/9 (66.7%)
       Dead7/17 (41.2%)4/8 (50.0%)3/9 (33.3%)
      Failure rates
       Local only2/18 (11.1%)0/8 (0.0%)2/10 (20.0%)
       Distant only2/18 (11.1%)2/8 (25.0%)0/10 (0.0%)
       Synchronous1/18 (5.6%)1/8 (12.5%)0/10 (0.0%)
      H&N, head and neck.
      a Indicates instance in which overall n = 17 owing to one patient receiving two discreet courses of treatment. When n = 17, de novo lesions n = 8 (47.0%), recurrent lesions n = 9 (53.0%).
      Nine of 18 (50.0%) tumors were adenocarcinomas (ACCs); the other nine of 18 (50.0%) were squamous cell carcinomas (SCCs). Ten of 18 esophageal or GEJ cancers (55.6%) represented locoregionally recurrent disease, whereas eight of 18 (44.4%) were de novo malignancies arising in a previously irradiated field. For the cohort of de novo tumors, previously irradiated malignancies included SCCs of the head and neck (37.5%), lung cancer (25.0%), Hodgkin’s lymphoma (25.0%), and thyroid diffuse large B-cell lymphoma (12.5%) (Table 1).
      The median cumulative RT dose delivered was 104.7 Gy (94.0–156.0 Gy), with a median time among treatment courses of 37.6 months (11.6–584 mo) (Table 2). There were two long-term interval outliers, with two patients having 584 month (48.6 y) and 450 month (37.5 y) intervals among the RT courses. Both outliers were treated remotely with mantle-field RT for Hodgkin’s lymphoma. Median re-RT dose per fraction was 1.8 Gy (1.1–2.0 Gy); seven of 18 of re-RT courses (38.8%) were delivered in a twice-daily (BID) fractionation in an attempt to minimize long-term toxicities (Table 2). The median re-RT dose was 53.4 Gy (40.0–108.0 Gy), with 108.0 Gy of re-RT being delivered to the patient who underwent three complete courses.
      Table 2RT Characteristics Stratified by De Novo Versus Recurrent Disease
      RT ParameterOverallDe Novo LesionsRecurrent Lesionsp Value
      Total dose—re-RT plan (Gy)
       Median (range)53.4 (40.8–61.2)56.7 (45.0–61.2)51.3 (48.0–59.4)0.170
      Total dose—previous RT plan (Gy)
       Median (range)50.4 (40.0–108.0
      Indicates instance in which overall n = 17 owing to one patient receiving two courses of re-RT (three total courses) for locally recidivistic disease. 108 Gy represents cumulative dose of first two plans. Third course of treatment given to a marginal recurrence with no area of the esophagus receiving full dose from all three courses.
      )
      50.4 (40.0–108.0
      Indicates instance in which overall n = 17 owing to one patient receiving two courses of re-RT (three total courses) for locally recidivistic disease. 108 Gy represents cumulative dose of first two plans. Third course of treatment given to a marginal recurrence with no area of the esophagus receiving full dose from all three courses.
      )
      50.4 (50.4–108.0
      Indicates instance in which overall n = 17 owing to one patient receiving two courses of re-RT (three total courses) for locally recidivistic disease. 108 Gy represents cumulative dose of first two plans. Third course of treatment given to a marginal recurrence with no area of the esophagus receiving full dose from all three courses.
      )
      0.893
      Cumulative RT dose (Gy)
       Median (range)104.7 (94.0–156.0)104.7 (94.0–134.8)103.8 (100.8–156.0)0.746
      Interval among treatment courses (mo)
       Median (range)37.6 (11.6–584.0)192 (18–584)31 (11.6–81.0)0.021
      Re-RT BID7/18 (38.9%)2/8 (25.0%)5/10 (50.0%)0.367
      Boldface indicates statistical significance (p < 0.05).
      Re-RT, re-irradiation; RT, irradiation.
      a Indicates instance in which overall n = 17 owing to one patient receiving two courses of re-RT (three total courses) for locally recidivistic disease. 108 Gy represents cumulative dose of first two plans. Third course of treatment given to a marginal recurrence with no area of the esophagus receiving full dose from all three courses.
      The patient who underwent a total of three courses of RT was initially treated to 50.4 Gy to the distal esophagus and regional nodes as part of trimodality therapy (2013) followed by 57.6 Gy to an anastomotic recurrence (2016) and an additional 48.0 Gy (2019, delivered in 1.2 Gy fractions BID) to slightly more proximal esophageal recurrence. When designing the RT fields for this third lesion, the inferior field edge was thought to be sufficiently superior to the upper border of the previous fields so that there would not be direct overlap of the upper airways and esophagus with this third course; however, proton RT was still recommended given the two previous exposures to other thoracic organs, such as the heart and lungs.
      Re-RT dose, initial RT dose, and cumulative RT dose did not differ among the treatment courses for de novo and recurrent disease; the only significant differences between de novo and recurrent esophageal + GEJ cancers seemed to be duration between courses of RT (median 31 [11.6–81.0] versus 192 [18–584] months, p = 0.0210) (Table 2).
      A total of 16 of 18 treatment courses (88.9%) were delivered with concurrent systemic therapy, which included carboplatin + paclitaxel (56.3%), capecitabine monotherapy (31.3%), capecitabine + cisplatin (6.3%), and pembrolizumab (6.3%) (Table 3). Of the two patients who did not receive concurrent systemic therapy, one patient was treated with RT alone for an inoperable de novo T1 lesion, and one patient was treated sequentially with full-dose cytotoxic chemotherapy preceding RT.
      Table 3Systemic Therapy Utilization
      Characteristicn = 18
      Concurrent systemic therapy
       Yes16/18 (88.9%)
       No2/18 (11.1%)
      Type of concurrent systemic therapy
       Carboplatin + paclitaxel9/16 (56.3%)
       Capecitabine5/16 (31.3%)
       Cisplatin + capecitabine1/16 (6.3%)
       Pembrolizumab1/16 (6.3%)
      Treatment was well tolerated, with only two incidences of G3 acute toxicities (dysphagia and esophagitis) and no acute G4 to G5 toxicities. Five of 18 treatment courses (27.8%) resulted in a G3 or higher late toxicity, with three of five events (3/18 courses, overall 16.7%) being G4 or higher. There were two G3 esophageal strictures requiring dilation, one incident each of G4 tracheal stenosis requiring permanent tracheostomy, and a tracheoesophageal fistula (TEF) requiring major surgical intervention. Of note, the patient diagnosed with tracheal stenosis was also dependent on percutaneous endoscopic gastrostomy tube feedings before re-RT, owing to dysphagia from his previous head and neck RT and the obstructive nature of his esophageal lesion. There was one G5 TEF leading to vertebral osteomyelitis and eventually, sepsis and death (Table 4). Given 15 of 18 plans included the proximal airways in the original and retreatment plans, there was an overall rate of TEF formation of 13.3%. All three G4 to G5 events occurred in the setting of a de novo esophageal SCC arising in a previously irradiated field. One of seven treatment courses delivered BID and four of 11 courses delivered once a day resulted in a G3 or higher late toxicity. The patient developing a G3 toxicity after BID RT had undergone three courses of RT and developed a G3 stricture requiring serial dilations after his third course of RT; however, this was likely due in part to local disease progression.
      Table 4Late Grade 3 or Higher Toxicities
      Toxicity (Grade)No. (%)Description
      Stenosis or strictures
       32 (11.1%)Esophageal strictures requiring dilation
       41 (5.6%)Tracheal stenosis requiring tracheostomy
      TEF
      Overall incidence of TEF when both courses of treatment involved the proximal airways (n = 15)
       41 (6.6%)TEF requiring surgical intervention
       51 (6.6%)TEF leading to vertebral osteomyelitis and death
      TEF, tracheoesophageal fistula.
      a Overall incidence of TEF when both courses of treatment involved the proximal airways (n = 15)
      With a median follow-up of 11.6 months (2.0–36.6), 10 of 17 (58.8%) of patients remained alive. Median overall survival was 19.5 months (95% confidence interval: 5.7–33.3); there was no difference in 1-year survival when stratified for primary versus recurrent disease (Table 5 and Figure 1—Supplemental Data 1). One-year LC was 75.3% plus or minus 17.4 and did not differ by primary versus recurrent disease (Table 5 and Figure 2—Supplemental Data 2). One-year DC was 83.3% plus or minus 11.2, and although it did not reach statistical significance, there was a trend toward worse DC with new primary disease as no patient who underwent treatment for a recurrent cancer experienced a distant failure (p = 0.117) (Table 5 and Figure 3—Supplemental Data 3).
      Table 5Median and 1-Year Oncologic Outcomes, Estimated Using the Kaplan-Meier Method
      OverallDe NovoRecurrentp Value
      Overall survival
       1 y68.9% ± 13.172.9% ± 16.562.5% ± 21.00.781
       Median (mo)19.5 (95% CI: 5.7–33.3)
      Local control
       1 y75.3% ± 17.466.7% ± 27.288.9% ± 10.50.507
       Median (mo)21.6 (95% CI: 7.2–36.3)
      Distant control
       1 y83.3% ± 11.272.9% ± 16.5100%0.117
       Median (mo)Not reached
      CI, confidence interval.
      All 18 PBS plans had comparative photon plans retrospectively generated. In one instance, a split-field intensity-modulated radiation therapy plan was generated to avoid a pacemaker; all other plans were VMAT based. Photon plans were optimized with the goal to match physician-drawn target coverage, while reducing OAR dose below physician-specified clinical constraints, or as low as achievable. When comparing coverage at 99%, 98%, and 95% of the CTVs, there were no differences in coverage observed among modalities (p = 0.84, 0.96, and 0.50, respectively). Doses to regional OARs, including the spinal cord (dMax, D0.1cc), total lung (mean, V20, V5), liver (for distal + GEJ tumors; mean), and heart (mean), were calculated for each plan and compared. All 18 plans evaluated the spinal cord and lungs; 16 of 18 (88.9%) evaluated the heart (two very proximal plans did not overlap the heart), and six of 18 plans (5.6%) with more distal targets evaluated the liver. Proton PBS plans were found to deliver significantly lower doses to the spinal cord, total lungs, liver, and heart (Table 6).
      Table 6Dosimetric Comparison of Radiation Dose Delivered to Critical OARs with PBS Versus Retrospectively Generated Optimized VMAT Plans Analyzed Using Wilcoxon Signed Rank Test
      Proton Median (Range)Photon Median (Range)Difference in Median (VMAT IMPT)Wilcoxon p Value
      Spinal cord
       DMax (Gy)13.1 (6.4–32.0)17.6 (8.6–39.9)4.50.04
       D0.1cc (Gy)10.5 (4.5–27.7)16.9 (8.0–33.4)6.40.01
      Total lung
       Mean (Gy)3.7 (0.39–7.10)6.9 (1.05–13.0)3.2<0.01
       V20 (%)5.4 (0.58–15.67)12.1 (0.61–27.5)6.7<0.01
       V5 (%)19.1(2.14–37.5)39.1 (4.8–64.1)20<0.01
      Liver
       Mean (Gy)0.3 (0.2–3.13)12.7 (0.01–19.3)12.50.02
      Heart
       Mean (Gy)0.70 (0–6.9)10.5 (.003–17.6)9.8<0.01
      Boldface indicates statistical significance (p < 0.05).
      OARs, organs at risk; PBS, pencil-beam scanning; VMAT, volumetric-modulated arc therapy; IMPT, intensity-modulated proton therapy.
      Prescription cumulative OAR dose constraints were variable and specified on the basis of physician discretion on an individualized basis; however, five of 18 VMAT plans (27.8%) would have been considered undeliverable owing to exceeding physician-specified cumulative OAR constraints; this included two plans for de novo and three plans for recurrent lesions. Physician-specified constraints for re-RT plans ranged from a spinal cord dMax of 4 to 35 Gy, with goals of achieving a cumulative dMax ranging from 44 to 65 Gy. For two of these five plans, the physician-specified spinal cord dose allowed by re-RT plans would have been exceeded with VMAT re-RT. In one instance, the re-RT PBS plans delivered a dMax of 8.4 Gy versus 23.0 Gy with VMAT (cumulative 49.8 versus 64.4 Gy), and in a second instance, the D0.1cc with PBS was 7.7 Gy versus 33.4 Gy with VMAT (cumulative 51.5 versus 77.2 Gy) with equivalent CTV coverage. Of two plans in which lung constraints would have been exceeded, in one patient, both the total lung mean in the VMAT versus the PBS plan was 6.1 Gy versus 13.0 Gy (cumulative 24.1 versus 31 Gy), and the total lung V20 was 11.3% versus 27.5% (cumulative 39.3% and 55.5%); in a second patient, the lung V20 was 15.7% versus 26.4% (cumulative 22.7% versus 28.7%) with VMAT. In a fifth patient, both the spinal cord dMax and the lung V20 were exceeded; PBS delivered a cord dMax of 13.4 versus 15.4 Gy (cumulative 52.4 versus 54.4 Gy) and a lung V20 of 12.2% versus 21.6% (cumulative 24.2% versus 33.6%).
      The median CTV volume was 135.4 cm3 (22.2–368.0). A total of 11 plans (61.1%) utilized single-field optimization intensity-modulated proton therapy (IMPT), meaning that each RT beam is optimized individually to deliver the prescribed dose, whereas eight plans (38.9%) used multifield optimization IMPT), meaning that all spots from all RT beams are optimized simultaneously (Table 7—Supplementary Data 4). The median number of beams used was three (range 2–4); beam geometries were variable and delineated on a case-by-case basis, but all plans using four beams underwent multifield optimization IMPT (Table 7—Supplementary Data 4). A total of 10 plans (55.6%) employed a range shifter which is utilized to draw PBS dose more superficially; range shifter depths ranged from 2 to 5 cm (Table 7—Supplementary Data 4). When generating CTVs, expansions were variable among treating physicians, but they generally ranged from 0.5 to 1.0 cm expansions on gross disease. Before CTV expansions, when applicable, an internal tumor volume expansion was generated to account for tumoral motion.
      Table 7RT Planning Parameters
      RT plan parametern = 18
      CTV volume (cm3)
       Median (range)135.4 (22.2–368.0)
      Optimization
       SFO11 (61.1%)
       MFO7 (38.9%)
      Number of beams
       26 (33.3%)
       37 (38.9%)
       45 (27.8%)
      Beam arrangements
       1 AO + 1 PO1 (5.6%)
       2 PO1 (5.6%)
       1 PA + 2 PO1 (5.6%)
       2 AO + 1 PA1 (5.6%)
       2 AO + 1 PO2 (11.1%)
       1 AP + 2 AO3 (16.7%)
       2 AO4 (22.2%)
       2 AO + 2 PO5 (27.8%)
      Range shifter
       None8 (44.4%)
       2 cm2 (11.1%)
       3 cm3 (16.7%)
       5 cm5 (27.8%)
      Elective nodal coverage
       Yes8 (44.4%)
       No10 (55.6%)
      Re-RT, re-irradiation; CTV, clinical target volume; SFO; single-field optimization; MFO, multifield optimization; AP, anterior-posterior; PA, posterior-anterior; AO, anterior oblique; PO, posterior oblique.
      No patient underwent surgical resection as management for their locoregional recurrences. A total of 15 of 18 treatments (83.3%) included portions of the proximal airways, including trachea, carina, or main stem bronchi, in both original and re-RT plans. Six of eight patients (75.0%) with a de novo malignancy presented with gross nodal disease in addition to the primary tumor, in comparison to four of 10 patients (40.0%) with recurrent disease (two nodal-only failures, two synchronous anastomotic + nodal failures). Three of eight (37.5%) patients with de novo malignancies were treated to gross primary and nodal disease with additional elective regional nodal coverage compared with five of 10 patients (50.0%) with recurrent disease. Overall, eight plans (44.4%) included elective nodal coverage in addition to the coverage of gross disease.
      The two patients with a remote history of classical mantle-field radiation for previous Hodgkin’s lymphoma did not have previous plans available for comparison given the number of years elapsed since treatment (delivered in 1970 and 1981). The prescription dose was assumed to be 40 Gy, standard for mantle-field RT, and given that in that era, treatment was delivered with one anterior-posterior and one posterior-anterior field, these patients were assumed to have received a previous dose of 40 to 45 Gy to the major thoracic organs, with the potential for “hotspots” as high as 110% of the prescription dose received by the spinal cord. In both cases, the new esophageal primaries developed directly within the high-dose region of a standard mantle-field plan, which would have typically covered the neck, mediastinum, and axillary nodal basins.

      Discussion

      We present a series of 17 patients who underwent 18 courses of re-RT with proton PBS for de novo or recurrent esophageal + GEJ malignancies. This is the largest series to date describing re-RT for these malignancies delivered with proton radiation and the first series that exclusively delivered re-RT with PBS. There are very limited data on re-RT for both de novo and recurrent esophageal + GEJ malignancies using definitive treatment approaches as presented here. This study describes favorable oncologic outcomes for a population with limited and poorly established treatment options and similar toxicity rates to other settings in which re-RT is utilized, including head and neck
      • Romesser P.B.
      • Cahlon O.
      • Scher E.D.
      • et al.
      Proton beam reirradiation for recurrent head and neck cancer: multi-institutional report on feasibility and early outcomes.
      ,
      • McDonald M.W.
      • Zolali-Meybodi O.
      • Lehnert S.J.
      • et al.
      Reirradiation of recurrent and second primary head and neck cancer with proton therapy.
      (referent) and lung
      • Romesser P.B.
      • Cahlon O.
      • Scher E.D.
      • et al.
      Proton beam reirradiation for recurrent head and neck cancer: multi-institutional report on feasibility and early outcomes.
      ,
      • McAvoy S.
      • Ciura K.
      • Wei C.
      • et al.
      Definitive reirradiation for locoregionally recurrent non-small cell lung cancer with proton beam therapy or intensity modulated radiation therapy: predictors of high-grade toxicity and survival outcomes.
      ,
      • Badiyan S.N.
      • Rutenberg M.S.
      • Hoppe B.S.
      • et al.
      Clinical outcomes of patients with recurrent lung cancer reirradiated with proton therapy on the Proton Collaborative Group and University of Florida Proton Therapy Institute Prospective registry studies.
      (referent).
      Although retrospective in nature, our series provides an important addition to this limited body of literature. Our reported rates of 27.8% G3 or higher late toxicities and 16.7% G4 or higher toxicities are consistent with to modestly improved in comparison to the limited historical comparators. Importantly, this came with encouraging rates of survival and disease control, with only three of 18 patients (16.7%) experiencing a local failure after re-RT. Interestingly, rates of local and DC at 1-year posttreatment were improved in the group of patients presenting with locally recurrent disease; and at the time of last follow-up, no patients in this cohort had developed distant metastatic disease. This may be in part owing to the biology of their recurrent disease and may suggest a specific pattern of behavior that could derive even more benefit from definitive local treatment.
      All de novo lesions were SCCs, although most of the recurrent lesions (9/10, 90.0%) were ACC. Although primary esophageal SCCs tend to be more likely to exhibit a locoregional pattern of recurrence as compared with distant hematogenous metastases,

      Barbetta A, Sihag S, Nobel T, et al. Patterns and risk of recurrence in patients with esophageal cancer with a pathologic complete response after chemoradiotherapy followed by surgery [e-pub ahead of print]. J Thorac Cardiovasc Surg. https://doi.org/10.1016/jtcvs.2018.09.136, accessed December 1, 2019.

      ,
      • Xi M.
      • Xu C.
      • Liao Z.
      • et al.
      The impact of histology on recurrence patterns in esophageal cancer treated with definitive chemoradiotherapy.
      there was a trend toward worse DC in our de novo cohort. This can potentially be explained by the hypothesis that many of these malignancies were likely RT-induced, and RT-induced malignancies tend to exhibit highly aggressive clinical courses, and that most of the de novo cases (75.0%) presented with gross nodal disease in addition to their primary esophageal lesions. RT-induced de novo esophageal cancers carry both local and distant recurrence risks at least as high as primary locally advanced esophageal cancers, whereas the patients in our recurrent cohort may represent a specific subset of disease with a more favorable pattern of distant metastases.
      Despite 1-year overall survival appearing worse in the patients with recurrent disease, it is important to note that two of three patients died from preexisting medical comorbidities. One patient died 8 months from re-RT from generalized failure to thrive and deconditioning, and one patient died 5 months from re-RT from dilated cardiomyopathy, attributed to long-standing ischemic cardiac disease present before both courses of RT. The third patient died with a third local recurrence but no distant metastatic disease over 3 years from his first local recurrence. Conversely, three of four deaths in the de novo malignancy cohort were from metastatic disease, and one of four was a treatment-related G5 toxicity. This further suggests that localized recurrences of esophageal + GEJ cancers should be managed aggressively and have the potential for long-term control.
      Despite optimal management with trimodality therapy, nearly 50% of patients with esophageal + GEJ cancer develop disease recurrence within 5 years.
      • Shapiro J.
      • van Lanschot JJB
      • Hulshof M.C.C.M.
      • et al.
      Neoadjuvant chemoradiotherapy plus surgery versus surgery alone for oesophageal or junctional cancer (CROSS): long-term results of a randomised controlled trial.
      Importantly, up to 22% of these failures have been described as locoregional failures (LRFs) only.
      • Shapiro J.
      • van Lanschot JJB
      • Hulshof M.C.C.M.
      • et al.
      Neoadjuvant chemoradiotherapy plus surgery versus surgery alone for oesophageal or junctional cancer (CROSS): long-term results of a randomised controlled trial.
      There are no consistent recommendations or guidelines on how to salvage LRFs, though surgical salvage and high-dose brachytherapy have both been investigated. Surgical salvage carries high rates of morbidity and mortality, with rates of anastomotic leakage and death ranging from 18.5% to 39% and 3% to 15%, respectively.
      • Markar S.R.
      • Karthikesalingam A.
      • Penna M.
      • Low D.E.
      Assessment of short-term clinical outcomes following salvage esophagectomy for the treatment of esophageal malignancy: systematic review and pooled analysis.
      • Swisher S.G.
      • Wynn P.
      • Putnam J.B.
      • et al.
      Salvage esophagectomy for recurrent tumors after definitive chemotherapy and radiotherapy.
      • Marks J.L.
      • Hofstetter W.
      • Correa A.M.
      • et al.
      Salvage esophagectomy after failed definitive chemoradiation for esophageal adenocarcinoma.
      • Miyata H.
      • Yamasaki M.
      • Takiguchi S.
      • et al.
      Salvage esophagectomy after definitive chemoradiotherapy for thoracic esophageal cancer.
      Combined analyses of high-dose brachytherapy for endoluminal lesions in medically inoperable patients or those with recurrent disease have reported lower mortality rates; however, this is a highly specialized treatment that is entirely dependent on a small cohort of specially trained individuals and can only be applied to small, localized endoluminal lesions.
      • Wong Hee Kam S.
      • Rivera S.
      • Hennequin C.
      • et al.
      Salvage high-dose-rate brachytherapy for esophageal cancer in previously irradiated patients: a retrospective analysis.
      • Nicolay N.H.
      • Rademacher J.
      • Oelmann-Avendano J.
      • Debus J.
      • Huber P.E.
      • Lindel K.
      High dose-rate endoluminal brachytherapy for primary and recurrent esophageal cancer: experience from a large single-center cohort.
      • Taggar A.S.
      • Pitter K.L.
      • Cohen G.N.
      • et al.
      Endoluminal high-dose-rate brachytherapy for locally recurrent or persistent esophageal cancer.
      In our series, 10 of 18 cases (55.6%), comprising six of eight (75.0%) de novo and four of 10 recurrent cases (40.0%), included gross nodal disease, which could not have been treated with brachytherapy alone. There is a great paucity of data regarding the role of re-RT with external-beam radiation in the management of recurrent esophageal cancers, and even less describing outcomes for de novo esophageal + GEJ cancers developing in fields of previous external-beam radiation, leaving this population of patients with markedly unmet oncologic needs.
      The role of proton RT in the management of esophageal + GEJ cancer is under active investigation. Retrospective series have reported improvements in survival, lymphopenia, and patient-reported quality of life outcomes for patients with locally advanced esophageal cancer treated with proton RT,
      • Xi M.
      • Xu C.
      • Liao Z.
      • et al.
      Comparative outcomes after definitive chemoradiotherapy using proton beam therapy versus intensity modulated radiation therapy for esophageal cancer: a retrospective, single-institutional analysis.
      • Fang P.
      • Shiraishi Y.
      • Verma V.
      • et al.
      Lymphocyte-sparing effect of proton therapy in patients with esophageal cancer treated with definitive chemoradiation.
      • Garant A.
      • Whitaker T.J.
      • Spears G.M.
      • et al.
      A comparison of patient-reported health-related quality of life during proton versus photon chemoradiation therapy for esophageal cancer.
      and recently, the results of the first prospective comparative modality trial have been presented and reported a decreased total toxicity burden when proton RT is utilized in the up-front setting for esophageal cancer without compromising oncologic outcomes.
      Comparing proton therapy to photon radiation therapy for esophageal cancer.
      ,
      • Lin S.H.
      • Hobbs B.
      • Thall P.
      • et al.
      Results of a phase II randomized trial of proton beam therapy vs intensity modulated radiation therapy in esophageal cancer.
      In addition, numerous dosimetric studies have revealed significant reductions in RT dose to the heart, lungs, and, in the case of distal esophageal tumors, liver when esophageal + GEJ cancer is treated with proton RT.
      • Hirano Y.
      • Onozawa M.
      • Hojo H.
      • et al.
      Dosimetric comparison between proton beam therapy and photon radiation therapy for locally advanced esophageal squamous cell carcinoma.
      • Shiraishi Y.
      • Xu C.
      • Yang J.
      • Komaki R.
      • Lin S.H.
      Dosimetric comparison to the heart and cardiac substructure in a large cohort of esophageal cancer patients treated with proton beam therapy or Intensity-modulated radiation therapy.
      • Welsh J.
      • Gomez D.
      • Palmer M.B.
      • et al.
      Intensity-modulated proton therapy further reduces normal tissue exposure during definitive therapy for locally advanced distal esophageal tumors: a dosimetric study.
      • Liu C.
      • Bhangoo R.S.
      • Sio T.T.
      • et al.
      Dosimetric comparison of distal esophageal carcinoma plans for patients treated with small-spot intensity-modulated proton versus volumetric-modulated arc therapies.
      These data are promising, and NRG (National Surgical Adjuvant Breast and Bowel Project/Radiation Therapy Oncology Group/Gynecologic Oncology Group) Oncology’s GI006 phase III randomized trial is underway. They warrant investigation of proton therapy in the re-RT setting as well.
      Proton radiation is a particularly attractive modality to utilize in the re-RT setting, particularly for esophageal + GEJ cancers. In recent years, prospective and large retrospective studies have evaluated the safety and efficacy of proton re-RT for recurrent head and neck, recurrent NSCLC, and recurrent soft tissue sarcomas, with results showing impressive rates of disease control, albeit with a 12% to 24.6% risk of G3 or higher late toxicities.
      • Romesser P.B.
      • Cahlon O.
      • Scher E.D.
      • et al.
      Proton beam reirradiation for recurrent head and neck cancer: multi-institutional report on feasibility and early outcomes.
      • Chao H.H.
      • Berman A.T.
      • Simone C.B.
      • et al.
      Multi-institutional prospective study of reirradiation with proton beam radiotherapy for locoregionally recurrent non-small cell lung cancer.
      • McDonald M.W.
      • Zolali-Meybodi O.
      • Lehnert S.J.
      • et al.
      Reirradiation of recurrent and second primary head and neck cancer with proton therapy.
      • McAvoy S.
      • Ciura K.
      • Wei C.
      • et al.
      Definitive reirradiation for locoregionally recurrent non-small cell lung cancer with proton beam therapy or intensity modulated radiation therapy: predictors of high-grade toxicity and survival outcomes.
      • Guttmann D.M.
      • Frick M.A.
      • Carmona R.
      • et al.
      A prospective study of proton reirradiation for recurrent and secondary soft tissue sarcoma.
      • Badiyan S.N.
      • Rutenberg M.S.
      • Hoppe B.S.
      • et al.
      Clinical outcomes of patients with recurrent lung cancer reirradiated with proton therapy on the Proton Collaborative Group and University of Florida Proton Therapy Institute Prospective registry studies.
      Most of our patients underwent treatment with a heterogeneous group of systemic therapies, given the mixed histologies of the included tumors and the number of lines of systemic therapy previously seen by many patients. Although the ability to draw conclusions about safety and efficacy of any specific combination is limited by sample size, physicians felt comfortable delivering re-RT concurrently based in part on extrapolations of previous thoracic and head and neck proton re-RT studies,
      • Romesser P.B.
      • Cahlon O.
      • Scher E.D.
      • et al.
      Proton beam reirradiation for recurrent head and neck cancer: multi-institutional report on feasibility and early outcomes.
      ,
      • Chao H.H.
      • Berman A.T.
      • Simone C.B.
      • et al.
      Multi-institutional prospective study of reirradiation with proton beam radiotherapy for locoregionally recurrent non-small cell lung cancer.
      ,
      • McAvoy S.
      • Ciura K.
      • Wei C.
      • et al.
      Definitive reirradiation for locoregionally recurrent non-small cell lung cancer with proton beam therapy or intensity modulated radiation therapy: predictors of high-grade toxicity and survival outcomes.
      and previous photon re-RT studies revealing improvements in lung cancer and toxicity profiles when increasingly conformal RT modalities were utilized.
      • Sulman E.P.
      • Schwartz D.L.
      • Le T.T.
      • et al.
      IMRT reirradiation of head and neck cancer-disease control and morbidity outcomes.
      ,
      • Sher D.J.
      • Haddad R.I.
      • Norris C.M.
      • et al.
      Efficacy and toxicity of reirradiation using intensity-modulated radiotherapy for recurrent or second primary head and neck cancer.
      Particularly in patients presenting with de novo, locally advanced malignancies, it was thought that chemotherapy or RT alone would unlikely be a curative treatment on the basis of previous randomized studies,
      • Cooper J.S.
      • Guo M.D.
      • Herskovic A.
      • et al.
      Chemoradiotherapy of locally advanced esophageal cancer: long-term follow-up of a prospective randomized trial (RTOG 85-01).
      and one of the largest prospective re-RT series to date has reported a survival benefit when re-RT was given with concurrent chemotherapy, though at the risk of increased acute toxicities.
      • Badiyan S.N.
      • Rutenberg M.S.
      • Hoppe B.S.
      • et al.
      Clinical outcomes of patients with recurrent lung cancer reirradiated with proton therapy on the Proton Collaborative Group and University of Florida Proton Therapy Institute Prospective registry studies.
      When designing RT fractionation schemes, fraction size and timing contribute to tumoral control and observed toxicities. When exposed to RT, most normal tissues can be categorized as “early” or “late” responding, based conceptually on their rate of cellular turnover and mathematically on a calculation known as the α-to-β ratio, which takes into account time, RT dose, and RT fractionation and is determined by the linear quadratic model for cellular survival curves. For example, gastrointestinal mucosal tissues tend to have a high turnover rate and a high α-to-β ratio and are early-responding tissues susceptible to acute toxicities. Major airways and nervous tissues, such as the spinal cord, have low turnover rates and low α-to-β ratios and are late-responding tissues susceptible to late toxicities. Tissues with low α-to-β ratios are thought to better tolerate “hyperfractionated” radiation, or radiation given in smaller doses more frequently, and by this rationale, when designing a re-RT plan, if cumulative dose to late-responding tissues such as the spinal cord or trachea are of greatest concern, physicians may opt to deliver RT in a hyperfractionated, BID fashion. When delivering re-RT to any area of the aerodigestive tract, late formation of strictures and fistulas become of paramount concern; when the spinal cord is re-irradiated, the risk of myelopathy is similarly troubling. When examining factors associated with the development with late toxicities within our series, no G4 or higher toxicity occurred in any patient who underwent BID treatment, delivered in fractions of 1.1 to 1.2 Gy. Though again limited by sample size, this warrants prospective investigation as a means to mitigate the risk of stricture and fistula formation.
      It is also worth noting that all three incidences of G4 to G5 late toxicities occurred in patients with de novo SCCs of the esophagus developing in a previous RT field; yet, all three patients were without evidence of local recurrence at the time of death or last follow-up. This is congruous with increased rates of pathologic complete response at the time of esophagectomy in SCC versus ACC
      • Shapiro J.
      • van Lanschot JJB
      • Hulshof M.C.C.M.
      • et al.
      Neoadjuvant chemoradiotherapy plus surgery versus surgery alone for oesophageal or junctional cancer (CROSS): long-term results of a randomised controlled trial.
      and is hypothesis generating, in that there could be consideration for future RT dose de-escalation for this particular subgroup. Although our numbers are small, there does not seem to be an association of G4 or higher late toxicity with type of concurrent systemic therapy. Of the three G4 or higher events, all three underwent different concurrent regimens (capecitabine, carbo + taxol, or capecitabine + cisplatin). In addition, there did not seem to be a correlation with neoadjuvant systemic therapy; of the three G4 or higher events, one patient had received neoadjuvant folinic acid + fluorouracil + oxaliplatin (FOLFOX); one, neoadjuvant carbo + taxol; and one, no neoadjuvant therapy.
      Clinical data for esophageal + GEJ re-RT in current literature remain limited. Photon radiotherapy has been reported as feasible, though only in small series and case reports, with high rates of late toxicities even when re-RT was delivered with palliative intent.
      • Katano A.
      • Yamashita H.
      • Nakagawa K.
      Re-irradiation of locoregional esophageal cancer recurrence following definitive chemoradiotherapy: a report of 6 cases.
      ,
      • Jingu K.
      • Niibe Y.
      • Yamashita H.
      • et al.
      Re-irradiation for oligo-recurrence from esophageal cancer with radiotherapy history: a multi-institutional study.
      ,
      • Sponseller P.
      • Lenards N.
      • Kusano A.
      • Patel S.
      Radiation treatment for newly diagnosed esophageal cancer with prior radiation to the thoracic cavity.
      ,
      • Yamaguchi S.
      • Ohguri T.
      • Imada H.
      • et al.
      Multimodal approaches including three-dimensional conformal re-irradiation for recurrent or persistent esophageal cancer: preliminary results.
      A contemporary propensity score-matched analysis was more promising and has suggested survival benefit from re-RT, which may reflect the increasing precision of RT in the modern era.
      • Hong L.
      • Huang Y.
      • Zhuang Q.
      • et al.
      Survival benefit of re-irradiation in esophageal cancer patients with locoregional recurrence: a propensity score-matched analysis.
      To date, series describing re-RT delivered with proton radiation are even more scarce and smaller,
      • Fernandes A.
      • Berman A.T.
      • Mick R.
      • et al.
      A prospective study of proton beam reirradiation for esophageal cancer.
      ,
      • Patel S.A.
      • Edgington S.K.
      • Adams J.
      • Morse C.
      • Ryan D.P.
      • Hong T.S.
      Novel use of proton beam therapy for neoadjuvant treatment of radiation-associated squamous cell carcinoma of the esophagus.
      ,
      • Berman A.T.
      • Fernandes A.T.
      • Both S.
      • et al.
      Prospective trial of proton reirradiation of locally recurrent esophageal cancer.
      and the only form of prospective data examining re-RT for locoregionally recurrent esophageal + GEJ cancers comes from an updated single-institutional analysis of 14 patients that revealed feasibility of proton re-RT delivery, but with a 35.7% rate of G3 or higher toxicities.
      • Fernandes A.
      • Berman A.T.
      • Mick R.
      • et al.
      A prospective study of proton beam reirradiation for esophageal cancer.
      ,
      • Berman A.T.
      • Fernandes A.T.
      • Both S.
      • et al.
      Prospective trial of proton reirradiation of locally recurrent esophageal cancer.
      Our in silico dosimetric comparison is in agreement with previously described dosimetric literature and supports the suggestion that when re-RT is undertaken, proton radiation should be strongly, if not exclusively, considered when available.
      • Zhang X.
      • Li Y.
      • Pan X.
      • et al.
      Intensity-modulated proton therapy reduces the dose to normal tissue compared with intensity-modulated radiation therapy or passive scattering proton therapy and enables individualized radical radiotherapy for extensive stage IIIB non-small-cell lung cancer: a virtual clinical study.
      • Lin L.
      • Kang M.
      • Huang S.
      • et al.
      Beam-specific planning target volumes incorporating 4D CT for pencil beam scanning proton therapy of thoracic tumors.
      • Verma V.
      • Rwigema J.C.M.
      • Malyapa R.S.
      • Regine W.F.
      • Simone C.B.
      Systematic assessment of clinical outcomes and toxicities of proton radiotherapy for reirradiation.
      The ability of proton RT, and particularly PBS, to modulate RT dose and have better conformity and dose falloff just distal to RT targets compared with other modalities used in all previous reports was likely a mitigating factor in our low observed rates of late toxicities. In our series, the differences in OAR dose exposure were most notable in the decreased mean liver dose (median 0.3 Gy versus 12.7 Gy, p = 0.02) and mean heart doses (median 0.70 Gy versus 10.5 Gy, p < 0.01), reflecting the remarkably limited low-dose spread of proton PBS RT. Given the risk factors that predispose for esophageal, lung, and head and neck malignancies, including obesity, tobacco use, and alcohol consumption, these patients are often already at risk of cardiopulmonary and hepatic diseases, and optimal sparing of these dose-limiting OARs will continue to be of paramount importance especially in the re-RT setting.
      This series contributes to a scarce body of literature for a very high-risk population. At the time of publication, systemic therapy for recurrent and de novo esophageal + GEJ malignancies previously treated with RT also remains poorly established, though investigations into the role of immunotherapy are actively underway.
      • Shah M.A.
      • Kojima T.
      • Hochhauser D.
      • et al.
      Efficacy and safety of pembrolizumab for heavily pretreated patients with advanced, metastatic adenocarcinoma or squamous cell carcinoma of the esophagus: the phase 2 KEYNOTE-180 study.
      • Fuchs C.S.
      • Doi T.
      • Jang R.W.
      • et al.
      Safety and efficacy of pembrolizumab monotherapy in patients with previously treated advanced gastric and gastroesophageal junction cancer: phase 2 clinical KEYNOTE-059 trial.

      Doi T, Piha-Paul SA, Jalal SI, et al. Safety and antitumor activity of the anti-programmed death-1 antibody pembrolizumab in patients with advanced esophageal carcinoma. In: J Clin Oncol. 2018;36:61–67.

      • Kato K.
      • Shah M.A.
      • Enzinger P.
      • et al.
      KEYNOTE-590: phase III study of first-line chemotherapy with or without pembrolizumab for advanced esophageal cancer.
      The results of this series and the previously published literature warrant ongoing prospective investigation into the role of proton RT in the re-RT setting for recurrent and de novo esophageal + GEJ malignancies, including novel immunotherapy-based regimens in combination with re-RT.

      Conclusions

      Re-RT for primary or recurrent malignancies of the esophagus + GEJ, when delivered with PBS proton therapy, yields high rates of LC and overall survival with acceptable toxicity profiles in a high-risk population.
      Proton PBS plans delivered significantly less radiation to critical OARs, including spinal cord (dMax, D0.1cc), total lungs (mean, V20, V5), and liver (mean) versus generated comparative VMAT photon plans.
      This is the largest series of re-RT for esophageal malignancies using proton RT and the first describing exclusive use of proton PBS; this also represents the first published series to date providing comparative dosimetric data between modalities in the re-RT setting. This study suggests re-RT with proton PBS as a feasible and well-tolerated treatment for this high-risk population of patients with limited treatment options and warrants ongoing prospective investigation, particularly with novel systemic agents, to further improve oncologic outcomes.

      References

      1. Surveillance, Epidemiology, and End Results (SEER) Program. 1973–2015, National Cancer Institute, DCCPS, Surveillance Research Program, released Cancer Stat Facts:2017.
        • Fernandes A.
        • Berman A.T.
        • Mick R.
        • et al.
        A prospective study of proton beam reirradiation for esophageal cancer.
        Int J Radiat Oncol. 2016; 95: 483-487
        • Katano A.
        • Yamashita H.
        • Nakagawa K.
        Re-irradiation of locoregional esophageal cancer recurrence following definitive chemoradiotherapy: a report of 6 cases.
        Mol Clin Oncol. 2017; 7: 681-686
        • Patel S.A.
        • Edgington S.K.
        • Adams J.
        • Morse C.
        • Ryan D.P.
        • Hong T.S.
        Novel use of proton beam therapy for neoadjuvant treatment of radiation-associated squamous cell carcinoma of the esophagus.
        J Gastrointest Oncol. 2019; 10: 155-160
        • Jingu K.
        • Niibe Y.
        • Yamashita H.
        • et al.
        Re-irradiation for oligo-recurrence from esophageal cancer with radiotherapy history: a multi-institutional study.
        Radiat Oncol. 2017; 12: 146
        • Kim Y.S.
        • Lee C.G.
        • Kim K.H.
        • et al.
        Re-irradiation of recurrent esophageal cancer after primary definitive radiotherapy.
        Radiat Oncol J. 2012; 30: 182-188
        • Hong L.
        • Huang Y.
        • Zhuang Q.
        • et al.
        Survival benefit of re-irradiation in esophageal cancer patients with locoregional recurrence: a propensity score-matched analysis.
        Radiat Oncol. 2018; 13: 171
        • Berman A.T.
        • Fernandes A.T.
        • Both S.
        • et al.
        Prospective trial of proton reirradiation of locally recurrent esophageal cancer.
        J Clin Oncol. 2014; 32 (159–159)
        • Romesser P.B.
        • Cahlon O.
        • Scher E.D.
        • et al.
        Proton beam reirradiation for recurrent head and neck cancer: multi-institutional report on feasibility and early outcomes.
        Int J Radiat Oncol Biol Phys. 2016; 95: 386-395
        • McDonald M.W.
        • Zolali-Meybodi O.
        • Lehnert S.J.
        • et al.
        Reirradiation of recurrent and second primary head and neck cancer with proton therapy.
        Int J Radiat Oncol Biol Phys. 2016; 96: 808-819
        • McAvoy S.
        • Ciura K.
        • Wei C.
        • et al.
        Definitive reirradiation for locoregionally recurrent non-small cell lung cancer with proton beam therapy or intensity modulated radiation therapy: predictors of high-grade toxicity and survival outcomes.
        Int J Radiat Oncol Biol Phys. 2014; 90: 819-827
        • Badiyan S.N.
        • Rutenberg M.S.
        • Hoppe B.S.
        • et al.
        Clinical outcomes of patients with recurrent lung cancer reirradiated with proton therapy on the Proton Collaborative Group and University of Florida Proton Therapy Institute Prospective registry studies.
        Pract Radiat Oncol. 2019; 9: 280-288
      2. Barbetta A, Sihag S, Nobel T, et al. Patterns and risk of recurrence in patients with esophageal cancer with a pathologic complete response after chemoradiotherapy followed by surgery [e-pub ahead of print]. J Thorac Cardiovasc Surg. https://doi.org/10.1016/jtcvs.2018.09.136, accessed December 1, 2019.

        • Xi M.
        • Xu C.
        • Liao Z.
        • et al.
        The impact of histology on recurrence patterns in esophageal cancer treated with definitive chemoradiotherapy.
        Radiother Oncol. 2017; 124: 318-324
        • Shapiro J.
        • van Lanschot JJB
        • Hulshof M.C.C.M.
        • et al.
        Neoadjuvant chemoradiotherapy plus surgery versus surgery alone for oesophageal or junctional cancer (CROSS): long-term results of a randomised controlled trial.
        Lancet Oncol. 2015; 16: 1090-1098
        • Markar S.R.
        • Karthikesalingam A.
        • Penna M.
        • Low D.E.
        Assessment of short-term clinical outcomes following salvage esophagectomy for the treatment of esophageal malignancy: systematic review and pooled analysis.
        Ann Surg Oncol. 2014; 21: 922-931
        • Swisher S.G.
        • Wynn P.
        • Putnam J.B.
        • et al.
        Salvage esophagectomy for recurrent tumors after definitive chemotherapy and radiotherapy.
        J Thorac Cardiovasc Surg. 2002; 123: 175-183
        • Marks J.L.
        • Hofstetter W.
        • Correa A.M.
        • et al.
        Salvage esophagectomy after failed definitive chemoradiation for esophageal adenocarcinoma.
        Ann Thorac Surg. 2012; 94 ([discussion: 1132–1133]): 1126-1132
        • Miyata H.
        • Yamasaki M.
        • Takiguchi S.
        • et al.
        Salvage esophagectomy after definitive chemoradiotherapy for thoracic esophageal cancer.
        J Surg Oncol. 2009; 100: 442-446
        • Wong Hee Kam S.
        • Rivera S.
        • Hennequin C.
        • et al.
        Salvage high-dose-rate brachytherapy for esophageal cancer in previously irradiated patients: a retrospective analysis.
        Brachytherapy. 2015; 14: 531-536
        • Nicolay N.H.
        • Rademacher J.
        • Oelmann-Avendano J.
        • Debus J.
        • Huber P.E.
        • Lindel K.
        High dose-rate endoluminal brachytherapy for primary and recurrent esophageal cancer: experience from a large single-center cohort.
        Strahlenther Onkol. 2016; 192: 458-466
        • Taggar A.S.
        • Pitter K.L.
        • Cohen G.N.
        • et al.
        Endoluminal high-dose-rate brachytherapy for locally recurrent or persistent esophageal cancer.
        Brachytherapy. 2018; 17: 621-627
        • Xi M.
        • Xu C.
        • Liao Z.
        • et al.
        Comparative outcomes after definitive chemoradiotherapy using proton beam therapy versus intensity modulated radiation therapy for esophageal cancer: a retrospective, single-institutional analysis.
        Int J Radiat Oncol Biol Phys. 2017; 99: 667-676
        • Fang P.
        • Shiraishi Y.
        • Verma V.
        • et al.
        Lymphocyte-sparing effect of proton therapy in patients with esophageal cancer treated with definitive chemoradiation.
        Int J Part Ther. 2018; 4: 23-32
        • Garant A.
        • Whitaker T.J.
        • Spears G.M.
        • et al.
        A comparison of patient-reported health-related quality of life during proton versus photon chemoradiation therapy for esophageal cancer.
        Pract Radiat Oncol. 2019; 9: 410-417
      3. Comparing proton therapy to photon radiation therapy for esophageal cancer.
        NCT03801876
        Date accessed: November 1, 2019
        • Lin S.H.
        • Hobbs B.
        • Thall P.
        • et al.
        Results of a phase II randomized trial of proton beam therapy vs intensity modulated radiation therapy in esophageal cancer.
        Int J Radiat Oncol. 2019; 105: 680-681
        • Hirano Y.
        • Onozawa M.
        • Hojo H.
        • et al.
        Dosimetric comparison between proton beam therapy and photon radiation therapy for locally advanced esophageal squamous cell carcinoma.
        Radiat Oncol. 2018; 13: 23
        • Shiraishi Y.
        • Xu C.
        • Yang J.
        • Komaki R.
        • Lin S.H.
        Dosimetric comparison to the heart and cardiac substructure in a large cohort of esophageal cancer patients treated with proton beam therapy or Intensity-modulated radiation therapy.
        Radiother Oncol. 2017; 125: 48-54
        • Welsh J.
        • Gomez D.
        • Palmer M.B.
        • et al.
        Intensity-modulated proton therapy further reduces normal tissue exposure during definitive therapy for locally advanced distal esophageal tumors: a dosimetric study.
        Int J Radiat Oncol Biol Phys. 2011; 81: 1336-1342
        • Liu C.
        • Bhangoo R.S.
        • Sio T.T.
        • et al.
        Dosimetric comparison of distal esophageal carcinoma plans for patients treated with small-spot intensity-modulated proton versus volumetric-modulated arc therapies.
        J Appl Clin Med Phys. 2019; 20: 15-27
        • Chao H.H.
        • Berman A.T.
        • Simone C.B.
        • et al.
        Multi-institutional prospective study of reirradiation with proton beam radiotherapy for locoregionally recurrent non-small cell lung cancer.
        J Thorac Oncol. 2017; 12: 281-292
        • Guttmann D.M.
        • Frick M.A.
        • Carmona R.
        • et al.
        A prospective study of proton reirradiation for recurrent and secondary soft tissue sarcoma.
        Radiother Oncol. 2017; 124: 271-276
        • Sulman E.P.
        • Schwartz D.L.
        • Le T.T.
        • et al.
        IMRT reirradiation of head and neck cancer-disease control and morbidity outcomes.
        Int J Radiat Oncol Biol Phys. 2009; 73: 399-409
        • Sher D.J.
        • Haddad R.I.
        • Norris C.M.
        • et al.
        Efficacy and toxicity of reirradiation using intensity-modulated radiotherapy for recurrent or second primary head and neck cancer.
        Cancer. 2010; 116: 4761-4768
        • Cooper J.S.
        • Guo M.D.
        • Herskovic A.
        • et al.
        Chemoradiotherapy of locally advanced esophageal cancer: long-term follow-up of a prospective randomized trial (RTOG 85-01).
        JAMA. 1999; 281: 1623-1627
        • Sponseller P.
        • Lenards N.
        • Kusano A.
        • Patel S.
        Radiation treatment for newly diagnosed esophageal cancer with prior radiation to the thoracic cavity.
        Med Dosim. 2014; 39: 251-255
        • Yamaguchi S.
        • Ohguri T.
        • Imada H.
        • et al.
        Multimodal approaches including three-dimensional conformal re-irradiation for recurrent or persistent esophageal cancer: preliminary results.
        J Radiat Res. 2011; 52: 812-820
        • Zhang X.
        • Li Y.
        • Pan X.
        • et al.
        Intensity-modulated proton therapy reduces the dose to normal tissue compared with intensity-modulated radiation therapy or passive scattering proton therapy and enables individualized radical radiotherapy for extensive stage IIIB non-small-cell lung cancer: a virtual clinical study.
        Int J Radiat Oncol Biol Phys. 2010; 77: 357-366
        • Lin L.
        • Kang M.
        • Huang S.
        • et al.
        Beam-specific planning target volumes incorporating 4D CT for pencil beam scanning proton therapy of thoracic tumors.
        J Appl Clin Med Phys. 2015; 16: 281-292
        • Verma V.
        • Rwigema J.C.M.
        • Malyapa R.S.
        • Regine W.F.
        • Simone C.B.
        Systematic assessment of clinical outcomes and toxicities of proton radiotherapy for reirradiation.
        Radiother Oncol. 2017; 125: 21-30
        • Shah M.A.
        • Kojima T.
        • Hochhauser D.
        • et al.
        Efficacy and safety of pembrolizumab for heavily pretreated patients with advanced, metastatic adenocarcinoma or squamous cell carcinoma of the esophagus: the phase 2 KEYNOTE-180 study.
        JAMA Oncol. 2019; 5: 546-550
        • Fuchs C.S.
        • Doi T.
        • Jang R.W.
        • et al.
        Safety and efficacy of pembrolizumab monotherapy in patients with previously treated advanced gastric and gastroesophageal junction cancer: phase 2 clinical KEYNOTE-059 trial.
        JAMA Oncol. 2018; 4e180013
      4. Doi T, Piha-Paul SA, Jalal SI, et al. Safety and antitumor activity of the anti-programmed death-1 antibody pembrolizumab in patients with advanced esophageal carcinoma. In: J Clin Oncol. 2018;36:61–67.

        • Kato K.
        • Shah M.A.
        • Enzinger P.
        • et al.
        KEYNOTE-590: phase III study of first-line chemotherapy with or without pembrolizumab for advanced esophageal cancer.
        Future Oncol. 2019; 15: 1057-1066