Abstract
Introduction:
While technical advances have allowed for reduced radiation doses to the lungs and heart when treating esophageal cancer (EC), doses to these organs cannot yet be eliminated completely. We conducted a post hoc analysis of CALGB 80803 (Alliance) to evaluate whether the radiation doses delivered to the heart and lungs in a prospective study with robust quality assurance were associated with overall survival (OS), and we assessed the impact of intensity modulated radiation therapy (IMRT) on adherence with protocol-specified dose constraints.
Methods:
Protocol-specified dose constraints were based on standard of care guidelines: total lung V10Gy ≤ 40%, V20Gy ≤ 20%, V30Gy ≤ 15%, V40Gy ≤ 10%; heart V30Gy ≤ 30%. Post hoc dose cutpoints were calculated that best differentiated OS.
Results:
There were 225 patients who underwent induction chemotherapy, where radiation modality data were available for 209/225 (92.8%); 127/209 (60.9%) were treated with IMRT. Adherence rates were high for each of the protocol-specified dose constraints: 162/209 (77.5%), 179/209 (85.6%), 207/209 (99.0%), 208/209 (99.5%), and 158/209 (75.6%); total lung: V10Gy, V20Gy, V30Gy, V40Gy, and heart V30Gy, respectively. The heart V30Gy ≤ 21% (p = 0.016) was associated with improved OS. Among patients with heart V30Gy ≤ 21%, 8.5% and 32.3% received 3-dimensional conformal radiotherapy (3D-CRT) and IMRT, respectively (p< 0.001). Mean heart V30Gy was lower in the IMRT versus the 3D-CRT group (24.7 versus 35.6; p< 0.001). On multivariable analysis, lung V10Gy >42% (hazard ratio [HR]: 1.71; p = 0.017) and heart V30Gy > 21% (HR: 1.86; p = 0.013) were associated with worse OS.
Conclusion:
Most patients on CALGB 80803 (Alliance) were treated with IMRT plans that met protocol-specified dose constraints for total lung and heart. Minimizing low dose volume to the lungs (V10Gy) and minimizing heart V30Gy were associated with a significant survival benefit. Total lung V10Gy ≤ 42% and heart V30Gy ≤ 21%, and utilization of IMRT over 3D-CRT, should be considered when designing radiation plans for EC patients.
INTRODUCTION
Esophageal cancer is an aggressive malignancy with a 5-year overall survival (OS) of 21% for all stages, and only 49% in patients with localized disease. Approximately 60% of newly diagnosed cases are adenocarcinomas.1 A standard of care in patients with locally advanced esophageal adenocarcinoma has been the use of trimodality therapy, consisting of neoadjuvant chemoradiation followed by esophagectomy, which has demonstrated an improvement in both OS and tumor control.2–4 Esophageal cancer was the sixth leading cause of cancer death in the United States in 2023,5,6 and current estimates indicate that the incidence is expected to increase dramatically over the next decade.7
While radiotherapy is an essential component of trimodality therapy, unintentional radiation dose to the lung and heart can be significant and even low doses to these organs are associated with higher risks of pulmonary and cardiac complications.8–11 However, studies that established these relationships were based on retrospective studies that largely utilized older, less conformal radiation techniques, such as three-dimensional conformal radiotherapy (3D-CRT).12,13 Hence, evaluating the impact of radiation dose to the heart and lung using more modern techniques, such as intensity modulated radiation therapy (IMRT), that allow for greater sparing of nearby organs at risk (OAR) is necessary to develop appropriate dose constraints. Moreover, it is critical to establish optimal strategies to minimize the risk of long-term treatment-related morbidity and mortality for patients undergoing therapy for esophageal cancer.
The Cancer and Leukemia Group B (CALGB)/Alliance 80803 trial was a randomized, phase 2 study designed to evaluate the early assessment of chemotherapy responsiveness by metabolic imaging to direct the choice of concurrent chemotherapy during chemoradiation in patients with esophageal cancer, with the goal of improving their response to therapy as demonstrated by pathologic complete response rates at the time of surgery. CALBG/Alliance 80803 was one of the first national cooperative group studies to incorporate IMRT planning for esophageal cancer into the protocol requirements, but the protocol also allowed for 3D CRT planning. Based on a priori knowledge from clinical experience, we anticipated that there may be differences in compliance with protocol-specified dose constraints between the planning methods. We conducted a post hoc analysis of this prospective clinical trial to: (1) evaluate compliance with protocol-specified total lung and heart radiation dose constraints; (2) determine the impact of the use of IMRT versus 3D-CRT on the ability to meet protocol-specified radiation dose constraints; and (3) assess the impact of total lung and heart radiation doses on OS.
MATERIALS AND METHODS
Study Design and Patients
The phase II randomized study was conducted by CALGB and enrolled patients at 69 outpatient cancer centers in the United States between November 9th 2011 and May 7th 2015.14 CALGB is now part of the Alliance for Clinical Trials in Oncology (Alliance). A total of 241 eligible patients with locally advanced esophageal and esophagogastric junction adenocarcinoma underwent a baseline PET scan and then were randomized to induction chemotherapy with either carboplatin/paclitaxel or folinic acid, 5-fluorouracil and oxaliplatin (mFOLFOX6). This was followed by a repeat PET scan and based on the PET response, patients continued on to chemoradiation with the same chemotherapy (responders) or switched to the alternative chemotherapy regimen (non-responders). Patients underwent surgery approximately 6-10 weeks after completion of chemoradiation. The study design, including eligibility criteria and treatment parameters on CALGB 80803 (Alliance) have been described previously.14
Dosimetric Data
Dose constraints specified in the study protocol were based on standard of care guidelines (total lung [V10Gy ≤ 40%, V20Gy ≤ 20%, V30Gy ≤ 15%, and V40Gy ≤ 10%] and heart [V30Gy ≤ 30%]).15–17 Dosimetric data were collected by the Quality Assurance Review Center (QARC). Discrete data elements, as well as individual data point from dose volume histograms, were extracted and analyzed.
Statistical Analysis
Patient characteristics by treatment modality (IMRT versus 3DCRT) were compared using the Pearson’s chi-squared test and the Kruskal-Wallis rank sum test for categorical and continuous variables, respectively. There were 217 patients with complete radiation dose data collected, including the dose constraints lung V10-V40Gy and heart V30Gy. Radiation dose to the lungs and heart was evaluated for these patients in relation to OS. Overall survival was measured from date of randomization until date of death from any cause. Relevant thresholds that best differentiated survival based on this radiation dose were evaluated using the “Survminer” R package;18 this method was used to algorithmically evaluate optimal cutpoints for each of the five dose constraints, based on a procedure proposed by Hothorn and Lausen.19 We then evaluated each binary radiation dose constraint in relation to OS using the Kaplan-Meier methods and log-rank tests.
Univariable and multivariable Cox proportional hazards models were used to evaluate the impact of demographic and treatment variables on OS. These results were presented with their associated hazard ratio (HR), 95% confidence interval (CI), and p-value, where the null hypothesis was rejected for p < 0.05.
The multivariable analysis began with a full model, which included the new dose constraints, and the following variables, based on clinical relevance: treatment received, response status (responder/non-responder), pathological complete response (pCR) combined with resection status, age, primary tumor site, race (white vs. other), histology grade, planning target volume (PTV), T category (T1/T2 vs. T3/T4), N category (AJCC 7th Edition), and patient sex. pCR with resection status was derived by combining resection classification (No resection, R0, and R1+) with pathologic complete response status (Yes, No). To obtain a final model to adjust for confounders of overall survival with respect to dose, backward stepwise regression procedure was then applied, in which variables were removed one at a time while evaluating model performance. Although the primary direction of selection was backward (i.e., removing variables), the procedure also allowed forward selection after the initial step, meaning that variables could be reintroduced if doing so improved the model performance. The criteria for model performance was based on the Akaike Information Criterion (AIC), a metric that balances model fit with parsimony (simpler models are preferred if they explain the data equally well), to avoid overfitting. Prior to model selection, the candidate variables were assessed for high multicollinearity using Spearman correlations. All comparisons had a correlation coefficient (rho) below an absolute value of 0.6. This approach ensured that the final model retained variables that contributed meaningfully to explaining overall survival. We recognize that step selection does not always find the globally optimal model. For the purposes of this study, we used it to explore whether any of the dosimetric cutpoints remained in the reduced model when weighed against other clinically relevant variables..
Data collection and statistical analyses were conducted by the Alliance Statistics and Data Management Center. Data quality was ensured by review of data by the Alliance Statistics and Data Management Center and by the study chairperson following Alliance policies. All analyses were based on the study database frozen on [07/30/2021].
RESULTS
Patient Demographic and Treatment Characteristics
Patient demographic and treatment characteristics by receipt of 3D-CRT and IMRT are presented in Table 1 (225 patients), which reflects patients who were eligible and had an interpretable PET scan after completing induction chemotherapy. The median follow-up was 5.2 years. There were 225 patients who underwent induction chemotherapy, 127 received IMRT, 82 received 3D-CRT, 8 received a combination of IMRT and 3D-CRT, and 8 did not go on to receive radiation therapy. The median age was 63 years, and 204/225 (90.7%) of patients were male. At baseline, the Eastern Cooperative Oncology Group (ECOG) performance status was 0 for 140/225 patients (62.2%) and 1 for the remaining 85/225 patients (37.8%). There were no significant differences in baseline patient characteristics between the patients receiving 3D-CRT and IMRT; IMRT was utilized in 127/225 (56.4%) of patients. Patients with available radiation modality data were treated across 65 centers, where 16, 32, and 17 utilized 3D-CRT only, IMRT only, and both 3D-CRT and IMRT, respectively.
Table 1.
Baseline Patient Characteristics (N = 225)
| Patient characteristic | 3D-CRT (n = 82) |
IMRT (n = 127) |
Other (n=16) |
p-value |
|---|---|---|---|---|
| Age (years) | 0.71 | |||
| Median | 63.3 | 62.6 | 63.1 | |
| Sex, No. (%) | 0.33 | |||
| Male | 73 (89.0%) | 118 (92.9%) | 13 (81.2%) | |
| Female | 9 (11.0%) | 9 (7.1%) | 3 (18.8%) | |
| Ethnicity, No. (%) | 0.86 | |||
| Non-Hispanic | 79 (96.3%) | 124 (97.6%) | 15 (93.8%) | |
| Hispanic or Latino | 2 (2.4%) | 2 (1.6%) | 0 (0.0%) | |
| Not Reported | 1 (1.2%) | 1 (0.8%) | 1 (6.2%) | |
| Race, No. (%) | 0.81 | |||
| White | 75 (91.5%) | 118 (92.9%) | 16 (100.0%) | |
| American Indian or Alaska Native | 1 (1.2%) | 1 (0.8%) | 0 (0.0%) | |
| Asian | 1 (1.2%) | 3 (2.4%) | 0 (0.0%) | |
| Black or African American | 3 (3.7%) | 3 (2.4%) | 0 (0.0%) | |
| Native Hawaiian or Pacific Islander | 0 (0.0%) | 1 (0.8%) | 0 (0.0%) | |
| Not Reported | 1 (1.2%) | 0 (0.0%) | 0 (0.0%) | |
| Unknown | 1 (1.2%) | 1 (0.8%) | 0 (0.0%) | |
| Histologic Grade (Differentiation), No. (%) | 0.75 | |||
| Well differentiated (Grade I) | 3 (3.7%) | 5 (4.0%) | 3 (18.8%) | |
| Moderately differentiated (Grade II) | 36 (43.9%) | 57 (45.2%) | 6 (37.5%) | |
| Poorly differentiated (Grade III) | 41 (50.0%) | 57 (45.2%) | 7 (43.8%) | |
| Undifferentiated (Grade IV) | 0 (0.0%) | 2 (1.6%) | 0 (0.0%) | |
| Unknown | 2 (2.4%) | 5 (4.0%) | 0 (0.0%) | |
| Missing | 0 | 1 | 0 | |
| ECOG Performance Status | 0.73 | |||
| 0 | 51 (62.2%) | 82 (64.6%) | 7 (43.8%) | |
| 1 | 31 (37.8%) | 45 (35.4%) | 9 (56.2%) | |
| T category (AJCC 7th Edition), No. (%) | 0.11 | |||
| T0 | 0 (0.0%) | 1 (0.8%) | 0 (0.0%) | |
| T1 | 4 (4.9%) | 1 (0.8%) | 0 (0.0%) | |
| T2 | 16 (19.5%) | 14 (11.1%) | 1 (6.2%) | |
| T3 | 62 (75.6%) | 107 (84.9%) | 14 (87.5%) | |
| T4 | 0 (0.0%) | 2 (1.6%) | 1 (6.2%) | |
| TX | 0 (0.0%) | 1 (0.8%) | 0 (0.0%) | |
| Missing | 0 | 1 | 0 | |
| N category (AJCC 7th Edition), No. (%) | 0.08 | |||
| N0 | 26 (31.7%) | 25 (19.8%) | 6 (37.5%) | |
| N1 | 35 (42.7%) | 74 (58.7%) | 6 (37.5%) | |
| N2 | 12 (14.6%) | 21 (16.7%) | 3 (18.8%) | |
| N3 | 6 (7.3%) | 3 (2.4%) | 0 (0.0%) | |
| NX | 3 (3.7%) | 3 (2.4%) | 1 (6.2%) | |
| Missing | 0 | 1 | 0 | |
| M category, No. (%) | 0.11 | |||
| M0 | 63 (76.8%) | 108 (85.7%) | 11 (68.8%) | |
| MX | 19 (23.2%) | 17 (13.5%) | 5 (31.2%) | |
| Unknown | 0 (0.0%) | 1 (0.8%) | 0 (0.0%) | |
| Missing | 0 | 1 | 0 | |
| Assigned Treatment, No (%) | 0.69 | |||
| Carboplatin / Paclitaxel | 43 (52.4%) | 63 (49.6%) | 8 (50.0%) | |
| mFOLFOX-6 | 39 (47.6%) | 64 (50.4%) | 8 (50.0%) | |
| Response Group, No. (%) | 0.51 | |||
| CP Non-Responder | 17 (20.7%) | 27 (21.3%) | 6 (37.5%) | |
| CP Responder | 26 (31.7%) | 36 (28.3%) | 2 (12.5%) | |
| FOLFOX Non-Responder | 17 (20.7%) | 19 (15.0%) | 3 (18.8%) | |
| FOLFOX Responder | 22 (26.8%) | 45 (35.4%) | 5 (31.2%) | |
| Resection Status, No. (%) | 0.97 | |||
| No resection | 13 (15.9%) | 21 (16.5%) | 7 (43.8%) | |
| Curative resection | 65 (79.3%) | 102 (80.3%) | 9 (56.2%) | |
| Palliative resection | 4 (4.9%) | 4 (3.1%) | 0 (0.0%) | |
| Type of Surgery, No. (%) | 0.39 | |||
| 3-hole esophagectomy (McKeown procedure) | 4 (4.9%) | 5 (3.94%) | 1 (6.3%) | |
| Ivor-Lewis esophagogastrectomy | 35 (42.6%) | 61 (48.0%) | 7 (43.8%) | |
| Minimally invasive esophagectomy | 13 (15.9%) | 17 (13.4%) | 1 (6.3%) | |
| Thoracoabdominal esophagectomy | 0 (0.0%) | 4 (3.15%) | 0 (0.0%) | |
| Transhiatal esophagectomy | 17 (20.7%) | 19 (15.0%) | 0 (0.0%) | |
| No Surgery | 13 (15.9%) | 21 (16.5%) | 7 (43.8%) | |
| Resection Margin Status, No. (%) | 0.97 | |||
| No Resection | 13 (15.9%) | 21 (16.5%) | 7 (43.8%) | |
| R0 (all gross tumor removed, microscopically negative margins) | 65 (79.3%) | 99 (78.0%) | 8 (50.0%) | |
| R1+ (microscopically positive margins or gross resedual tumor remains) | 4 (4.9%) | 7 (5.5%) | 1 (6.2%) | |
| Tumor Depth, No. (%) | 0.10 | |||
| No evidence of tumor at time of surgery | 20 (29.4%) | 35 (32.7%) | 6 (66.7%) | |
| Tumor directly invades adjacent structures | 1 (1.5%) | 2 (1.9%) | 0 (0.0%) | |
| Tumor invades adventitia | 8 (11.8%) | 28 (26.2%) | 2 (22.2%) | |
| Tumor invades lamina propria or submucosa | 16 (23.5%) | 20 (18.7%) | 0 (0.0%) | |
| Tumor invades muscularis propria | 23 (33.8%) | 22 (20.6%) | 1 (11.1%) | |
| Missing | 14 | 20 | 7 | |
| Responder, No. (%) | 0.45 | |||
| Non-Responder | 34 (41.5%) | 46 (36.2%) | 9 (56.2%) | |
| Responder | 48 (58.5%) | 81 (63.8%) | 7 (43.8%) | |
| Pathologic Complete Response (pCR), No. (%) | 0.68 | |||
| No | 64 (78.0%) | 96 (75.6%) | 10 (62.5%) | |
| Yes | 18 (22.0%) | 31 (24.4%) | 6 (37.5%) | |
| pCR with Resection Classification, No. (%) | 0.96 | |||
| No Resection and no pCR | 13 (15.9%) | 21 (16.5%) | 7 (43.8%) | |
| R0 and no pCR | 47 (57.3%) | 68 (53.5%) | 2 (12.5%) | |
| R0 with pCR | 18 (22.0%) | 31 (24.4%) | 6 (37.5%) | |
| R1 and no pCR | 4 (4.9%) | 7 (5.5%) | 1 (6.2%) | |
| Primary Tumor Site, No. (%) | 0.21 | |||
| Gastro-esophageal junction | 44 (53.7%) | 83 (65.9%) | 6 (37.5%) | |
| Lower thoracic esophagus (excludes G-E junction) | 28 (34.1%) | 31 (24.6%) | 5 (31.2%) | |
| Other | 10 (12.2%) | 12 (9.5%) | 5 (31.2%) | |
| Missing | 0 | 1 | 0 | |
| Planning Target Volume (cc) | 0.74 | |||
| Median | 714.6 | 689.5 | 768.4 | |
| Missing | 0 | 0 | 8 | |
Abbreviations: 3D-CRT: three-dimensional conformation radiation therapy; CP: carboplatin/paclitaxel; ECOG: Eastern Cooperative Oncology Group; FOLFOX: folinic acid, 5-FU, oxaliplatin; IMRT: intensity modulated radiation therapy; No.: number; R0: margin negative resection; R1: microscopically positive surgical margins
P-values in table 1 were calculated between 3D-CRT and IMRT excluding patients included in Other. The other column consisted of patients who received both IMRT and 3D-CRT (n=8) and those who did not undergo radiation treatment (n=8). P-values for comparisons between radiation therapy type and categorical variables were calculated using Pearson’s Chi-squared test. Additionally, p-values for numeric variables were calculated using Kruskal-Wallis rank sum test.
Following induction chemotherapy, 217 patients underwent radiation treatment (IMRT [n = 127], 3D-CRT [n = 82], and both [n = 8]). Each participant signed an IRB-approved, protocol-specific informed consent document in accordance with federal and institutional guidelines.
Compliance with Protocol-specified Total Lung and Heart Dose Constraints
Of the 217 patients with dose information, 209 had adequate radiation treatment modality information (received either 3D or IMRT exclusively). Overall compliance rates, as well as compliance by treatment modality for each of the protocol-specified dose constraints are presented in Table 2.
Table 2.
Protocol-specified Dose Constraints Compliance and by Radiation Modality (n = 209)
| Organ | Protocol-Specified Dose Constraint | Compliance | 3D-CRT | IMRT | D(%) [95% CI (%)] | p-value1 |
|---|---|---|---|---|---|---|
| Total Lung | V10 Gy ≤ 40% | 162/209 (77.5%) |
67/82 (81.7%) |
95/127 (74.8 %) |
6.9 [−4.9, 17.6] | 0.24 |
| V20 Gy ≤ 20% | 179/209 (85.6%) |
66/82 (80.5%) |
113/127 (89.0%) |
8.5 [−1.2, 19.3] | 0.09 | |
| V30 Gy ≤ 15% | 207/209 (99.0%) |
81/82 (98.8%) |
126/127 (99.2%) |
0.4 [−3.3, 5.8] | 0.75 | |
| V40 Gy ≤ 10% | 208/209 (99.5%) |
81/82 (98.8%) |
127/127 (100%) |
1.2 [−1.9, 6.6] | 0.21 | |
| Heart | V30 Gy ≤ 30% | 158/209 (75.6%) |
49/82 (59.8%) |
109/127 (85.8%) |
26.0 [13.8, 37.9] | < 0.001 |
Abbreviations: Vx: volume receiving at least X Gy of radiation
P-value calculated with Pearson Chi-Square test between 3D-CRT and IMRT comparing patients who met vs. those who didn’t meet protocol-specified dose constraint for each specification level.
Compliance rates with total lung dose constraints of V10Gy ≤ 40%, V20Gy ≤ 20%, V30Gy ≤ 15%, and V40Gy ≤ 10% were 162/209 (77.5%), 179/209 (85.6%), 207/209 (99.0%), and 208/209 (99.5%), respectively. Compliance with the heart V30Gy ≤ 30% dose constraint was 158/209 (75.6%). When evaluating compliance with protocol-specified dose constraints by radiation modality, there were no statistically significant differences in compliance rates for the total lung dose constraints by receipt of 3D-CRT or IMRT. However, rates of compliance with the heart V30Gy ≤ 30% dose constraint were significantly better with IMRT versus 3D-CRT (109/127 [85.8%] versus 49/82 [59.8%], respectively, difference of 26%, 95% CI [13.8%, 37.9%] [p < 0.001]).
Impact of Heart and Lung Doses on Overall Survival
Among 217 patients with radiation information, we calculated cutpoints that best differentiated OS for each of the total lung and heart dose constraints. For the total lung, these cutpoints were: V10Gy ≤ 42%, V20Gy ≤ 17%, V30Gy ≤ 8%, and V40Gy ≤ 4% (Figure 1A–D). The ability to achieve a total lung V40Gy ≤ 4% was significantly higher in patients who received IMRT 111/127 (87.4%) compared to 3D-CRT 62/82 (75.6%), p = 0.03 (Supplementary Table 1). For the heart dose constraint, the calculated cutpoint was V30Gy ≤ 21%. Use of IMRT (32.3% of patients) over 3D-CRT (8.5% of patients) was associated with a statistically significant advantage in being able to meet the heart V30Gy ≤ 21% (p < 0.001). These results are presented in Supplementary Table 1 and Figure 2A–B. Furthermore, the mean heart V30Gy was 24.7%, 95% CI [23.1%, 26.3%] versus 35.6%, 95% CI [32.4%, 38.8%] for the IMRT and 3D-CRT groups, respectively (p < 0.001), as shown in Figure 2B.
Figure 1. Impact of Total Lung Doses on Overall Survival.
A: Overall survival Kaplan-Meier plots for patients with a total lung V10 Gy ≤42% (n = 174; red curve) and total lung V10 Gy > 42% (n = 43; teal curve). The median OS was 55.7, 95% CI [34.5, NA] versus 21.7, 95% CI [17.5, 47.4] months for the V10 Gy ≤42% and V10 Gy > 42% groups, respectively (p = 0.004).
B: Overall survival Kaplan-Meier plots for patients with a total lung V20 Gy ≤17% (n = 135; red curve) and total lung V20 Gy > 17% (n = 82; teal curve). The median OS was 67.4, 95% CI [34, NA] versus 32.8, 95% CI [21.7, 47.3] months for the V20 Gy ≤17% and V20 Gy > 17% groups, respectively (p = 0.018).
C: Overall survival Kaplan-Meier plots for patients with a total lung V30 Gy ≤8% (n = 166; red curve) and total lung V30 Gy > 8% (n = 51; teal curve). The median OS was 52.4, 95% CI [34, NA] versus 27.3, 95% CI [18.9, NA] months for the V30 Gy ≤8% and V20 Gy > 8% groups, respectively (p = 0.032).
D: Overall survival Kaplan-Meier plots for patients with a total lung V40 Gy ≤4% (n = 177; red curve) and total lung V40 Gy > 4% (n = 40; teal curve). The median OS was 51.1, 95% CI [34, NA] versus 27.8. 95% CI [18, NA] months for the V40 Gy ≤4% and V40 Gy > 4% groups, respectively (p = 0.025).
Abbreviations: Gy: gray; n: number of patients; OS: overall survival
Figure 2. Impact of Heart Dose on Overall Survival.
A: Overall survival Kaplan-Meier plots for 217 patients with a heart V30 Gy ≤21% (n = 168; red curve) and heart V30 Gy > 21% (n = 49; teal curve). The median OS was 73.9, 95% CI [52.4, NA] versus 34.5, 95% CI [27.8, 51.1] months for the V30 Gy ≤21% and V30 Gy > 21% groups, respectively (p = 0.016).
B: Box and whisker plot of heart V30 Gy doses in patients receiving 3D-CRT or IMRT. The mean heart V30 Gy doses were 35.6%, 95% CI [32.5, 38.7] versus 24.7%, 95% CI [23.2, 26.2] for the 3D-CRT and IMRT groups, respectively (p < 0.001).
The median OS for all patients with a total lung V10Gy ≤ 42% versus > 42% was 55.7 months, 95% CI [34.5, not reached] versus 21.7 months, 95% CI [17.5, 47.4], respectively (p = 0.004). In patients with a total lung V20Gy ≤ 17% versus > 17%, the median OS was 67.4 months, 95% CI [34.0, not reached] versus 32.8 months, 95% CI [21.7, 47.3], respectively (p = 0.018). In patients with a total lung V30Gy ≤ 8% versus > 8%, the median OS was 52.4 months, 95% CI [34.0, not reached] versus 27.3 months, 95% CI [18.9, not reached], respectively (p = 0.032). In patients with a total lung V40Gy ≤ 4% versus > 4%, the median OS was 51.1 months, 95% CI [34.0, not reached] versus 27.8 months, 95% CI [18.0, not reached], respectively (p = 0.025). These results are presented in Table 3 and Figure 1A–D. The median OS in patients with a heart V30Gy ≤ 21% versus > 21% was 73.9 months, 95% CI [52.4, not reached] versus 34.5 months, 95% CI [27.8, 51.1], respectively (p = 0.016), as shown in Table 3 and Figure 2A.
Table 3.
Median Overall Survival by Calculated Dosimetric Cutpoints
| Organ | Calculated Dose Constraint | Number of Patients | Median Overall Survival (months) [95%CI] | p-value |
|---|---|---|---|---|
| Total Lung | V10 Gy ≤42% V10 Gy > 42% |
174 43 |
55.7 [34.5, NR] 21.7 [17.5, 47.4] |
0.004 |
| V20 Gy ≤17% V20 Gy > 17% |
135 82 |
67.4 [34.0, NR] 32.8 [21.7, 47.3] |
0.018 | |
| V30 Gy ≤8% V30 Gy > 8% |
166 51 |
52.4 [34.0, NR] 27.3 [18.9, NR] |
0.032 | |
| V40 Gy ≤4% V40 Gy > 4% |
177 40 |
51.1 [34.0, NR] 27.8 [18.0, NR] |
0.025 | |
| Heart | V30 Gy ≤21% V30 Gy > 21% |
49 168 |
73.9 [52.4, NR] 34.5 [27.8, 51.1] |
0.016 |
Abbreviations: Vx: volume receiving at least X Gy of radiation
NR: Not reached due to 5-year maximum follow-up for OS
Prognostic Factors for Overall Survival
Univariable and multivariable analyses evaluating prognostic factors for OS are presented in Table 4. On univariable analysis, factors prognostic for worse OS were: total lung V10Gy > 42% (p = 0.002), total lung V10Gy > 40% (p = 0.03), total lung V20Gy > 17% (p = 0.019), total lung V20Gy > 20% (p = 0.042), total lung V30Gy > 8% (p = 0.033), total lung V40 > 4% (p = 0.027), and heart V30Gy > 21% (p = 0.011), pT1-2 category (p < 0.001), pT3-4 category (p < 0.001), pN1-pNX category (pN1 (p = 0.034), pN2 (p = 0.02), pN3 = (p = 0.002), pNX (p = 0.026), ECOG performance status (p = 0.015), and increasing PTV volume (p = 0.005). Achieving an R0 resection (p < 0.001) or an R1 resection (p = 0.005) was prognostic for improved OS on univariable analysis.
Table 4.
Univariable and Multivariable Cox Regression for Overall Survival
| Univariable Analysis | ||||||
|---|---|---|---|---|---|---|
| Variable | Hazard Ratio | 95 % CI | p-value | |||
| Ethnicity | ||||||
| Hispanic | Ref | Ref | Ref | |||
| Non-Hispanic | 1.35 | 0.33-5.45 | 0.7 | |||
| Not Reported | 0.83 | 0.08-9.16 | 0.9 | |||
| Total Lung V10 Gy (Cutpoint) | Multivariate | |||||
| V10 Gy ≤42% | Ref | Ref | Ref | Hazard Ratio | 95% CI | p-value |
| V10 Gy > 42% | 1.88 | 1.25-2.83 | 0.002 | 1.71 | 1.10-2.66 | 0.017 |
| Total Lung V10 Gy (Protocol) | ||||||
| V10 Gy ≤40% | Ref | Ref | Ref | |||
| V10 Gy > 40% | 1.55 | 1.04-2.30 | 0.03 | |||
| Total Lung V20 (Cutpoint) | ||||||
| V20 Gy ≤17% | Ref | Ref | Ref | |||
| V20 Gy > 17% | 1.55 | 1.08-2.23 | 0.019 | |||
| Total Lung V20 Gy (Protocol) | ||||||
| V20 Gy ≤20% | Ref | Ref | Ref | |||
| V20 Gy > 20% | 1.61 | 1.02-2.54 | 0.042 | |||
| Total Lung V30 Gy (Cutpoint) | ||||||
| V30 Gy ≤8% | Ref | Ref | Ref | |||
| V30 Gy > 8% | 1.55 | 1.04-2.32 | 0.033 | |||
| Total Lung V30 Gy (Protocol) | ||||||
| V30 Gy ≤15% | Ref | Ref | Ref | |||
| V30 Gy > 15% | 1.23 | 0.30-5.00 | 0.8 | |||
| Total Lung V40 Gy (Cutpoint) | ||||||
| V40 Gy ≤4% | Ref | Ref | Ref | |||
| V40 Gy > 4% | 1.63 | 1.06-2.50 | 0.027 | |||
| Total Lung V40 Gy (Protocol) | ||||||
| V40 Gy ≤10% | Ref | Ref | Ref | |||
| V40 Gy > 10% | 0.66 | 0.09-4.76 | 0.7 | |||
| Heart V30 Gy (Cutpoint) | Multivariate | |||||
| V30 Gy ≤21% | Ref | Ref | Ref | Hazard Ratio | 95% CI | p-value |
| V30 Gy > 21% | 1.87 | 1.16-3.03 | 0.011 | 1.86 | 1.14-3.05 | 0.013 |
| Heart V30 Gy (Protocol) | ||||||
| V30 Gy ≤30% | Ref | Ref | Ref | |||
| V30 Gy > 30% | 1.40 | 0.94, 2.08 | 0.10 | |||
| Assigned Chemotherapy | ||||||
| Carboplatin/Paclitaxel | Ref | Ref | Ref | |||
| mFOLFOX | 0.80 | 0.56-1.15 | 0.2 | |||
| Response Group | ||||||
| Carboplatin/Paclitaxel Responder | Ref | Ref | Ref | |||
| Carboplatin/Paclitaxel Non-Responder | 1.09 | 0.67-1.78 | 0.7 | |||
| FOLFOX Non-Responder | 1.14 | 0.68-1.91 | 0.6 | |||
| FOLFOX Responder | 0.70 | 0.44-1.11 | 0.13 | |||
| Resection Status | ||||||
| No resection | Ref | Ref | Ref | |||
| Curative resection | 0.14 | 0.09-0.22 | <0.001 | |||
| Palliative resection | 0.20 | 0.07-0.56 | 0.002 | |||
| Resection Class | ||||||
| No Resection | Ref | Ref | Ref | |||
| R0 (all gross tumor removed, microscopically negative margins) | 0.13 | 0.09-0.20 | <0.001 | |||
| R1+ (microscopically positive margins or gross residual tumor remains) | 0.36 | 0.18-0.74 | 0.005 | |||
| Pathologic Complete Response with Resection Class | Multivariate | |||||
| No resection and no pCR | Ref | Ref | Ref | Hazard Ratio | 95% CI | p-value |
| R0 and no pCR | 0.18 | 0.12-0.27 | <0.001 | 0.168 | 0.11-0.27 | <0.001 |
| R0 and pCR | 0.05 | 0.03-0.11 | <0.001 | 0.05 | 0.02-0.09 | <0.001 |
| R1+ and no pCR | 0.36 | 0.18-0.73 | 0.005 | 0.38 | 0.18-0.79 | 0.009 |
| Tumor Depth | ||||||
| No evidence of tumor at time of surgery | Ref | Ref | Ref | |||
| T1(Tumor invades lamina propria or submucosa)/T2(Tumor invades muscularis propria) | 2.88 | 1.57-5.27 | <0.001 | |||
| T3 (Tumor invades adventitia)/ T4 (Tumor directly invades adjacent structures) | 4.59 | 2.40-8.78 | <0.001 | |||
| Pathologic Complete Response | ||||||
| No | Ref | Ref | Ref | |||
| Yes | 0.22 | 0.12-0.39 | <0.001 | |||
| Primary Site | ||||||
| Gastroesophageal Junction | Ref | Ref | Ref | |||
| Lower Thoracic Esophagus | 1.03 | 0.69-1.55 | 0.9 | |||
| Other | 1.01 | 0.58-1.76 | >0.9 | |||
| Radiation Modality | ||||||
| 3D-CRT | Ref | Ref | Ref | |||
| IMRT | 0.97 | 0.67-1.41 | 0.9 | |||
| Both | 0.58 | 0.18-1.87 | 0.4 | |||
| Patient Sex | ||||||
| Female | Ref | Ref | Ref | |||
| Male | 1.27 | 0.64-2.50 | 0.5 | |||
| Patient Race | Multivariate | |||||
| Other | Ref | Ref | Ref | Hazard Ratio | 95% CI | p-value |
| White | 1.10 | 0.54-2.26 | 0.8 | 1.79 | 0.84-3.81 | 0.132 |
| Histologic Grade | ||||||
| Well Differentiated (Grade I) | Ref | Ref | Ref | |||
| Moderately Differentiated (Grade II) | 0.72 | 0.33-1.58 | 0.4 | |||
| Poorly Differentiated (Grade III) | 0.64 | 0.29-1.41 | 0.3 | |||
| Undifferentiated (Grade IV) | 1.75 | 0.36-8.47 | 0.5 | |||
| Unknown | 0.31 | 0.06-1.48 | 0.14 | |||
| pN category | ||||||
| N0 | Ref | Ref | Ref | |||
| N1 | 1.70 | 1.04-2.79 | 0.034 | |||
| N2 | 2.02 | 1.12-3.66 | 0.020 | |||
| N3 | 4.18 | 1.68-10.5 | 0.002 | |||
| NX | 3.04 | 1.15-8.09 | 0.026 | |||
| Age (Years) | 1.01 | 0.99-1.03 | 0.3 | |||
| ECOG Performance Status | 1.56 | 1.09-2.24 | 0.015 | |||
| Planning Target Volume (100 cm3) | 1.10 | 1.03-1.18 | 0.005 | |||
Abbreviations: CI: confidence interval; CP: carboplatin/paclitaxel;; ECOG: Eastern Cooperative Oncology Group; FOLFOX: folinic acid, 5-FU, oxaliplatin; R1: surgical resection with microscopic positive margins; R2: surgical resection with gross positive margins; Ref: reference; Vx: volume receiving at least X Gy of radiation
Multivariable analysis after step selection retained: total lung V10Gy > 42%, heart V30Gy > 21%, a composite variable of pCR with resection classification, and binary race status (White vs. Other). Negative prognostic factors for OS were total lung V10 > 42% (p = 0.017) and heart V30Gy > 21% (p = 0.013). All three levels for the composite variable of pCR with resection classification were prognostic for improved OS when compared to the baseline of no surgery/no pCR. Achieving a pCR with a resection classification of R0 demonstrated the lowest hazard (p < 0.001), followed by achieving a resection classification of R0 with no pCR (p < 0.001), and lastly achieving a resection classification of R1 with no pCR (p = 0.009).
DISCUSSION
Compliance with protocol-specified dose constraints to the total lung and heart was excellent on the CALGB 80803 study; over 70% of radiation treatment plans met each of the protocol-specified constraints. This demonstrated the benefit of treating patients on clinical trials, where there are clear radiation therapy guidelines and quality assurance procedures in place to ensure that patients receive high quality care. We also evaluated the impact of IMRT over 3D-CRT on the ability to meet these dose constraints. IMRT, as opposed to 3D-CRT, allowed for a significantly greater ability to meet the heart constraint. Furthermore, we identified novel dose constraints that demonstrated an improvement in OS and were generally easier to meet with IMRT than with 3D-CRT. To our knowledge, this is the first prospective trial reporting on the impact of lung and heart dose on clinical outcomes.
Other groups have also demonstrated a benefit of IMRT versus 3D-CRT in terms of reducing mortality after chemoradiation for esophageal cancer, in particular from cardiac causes. In 2012, Lin et al. conducted a single institution retrospective analysis of 676 patients with esophageal cancer who underwent chemoradiotherapy for esophageal cancer.13 They observed that the use of IMRT over 3D-CRT was associated with a significant improvement in OS (HR: 0.72; p < 0.001). These investigators observed a higher incidence of documented cardiac-related deaths with 3D-CRT versus IMRT, likely related to less sparing of the heart with 3D-CRT. 13 Similarly, we observed that the use of IMRT over 3D-CRT was associated with a significantly improved ability to meet the protocol-specified (V30Gy ≤ 30%) and calculated (V30Gy ≤ 21%) heart dose constraints, as shown in Table 2 and Supplementary Table 1. This study adds to their findings by assessing the impact of IMRT in a prospective dataset and providing a dose constraint for the heart that is associated with improved OS and more readily achievable with IMRT over 3D-CRT (V30Gy ≤ 21%), as shown in Figure 2.
A multi-institutional study published by Garant et al. in 2021 evaluated the impact of radiation treatment modality on dosimetric predictors of cardiopulmonary toxicities following trimodality therapy for esophageal cancer.10 This was a retrospective analysis of 465 patients who received 3D-CRT (n = 178), IMRT (n = 192), and proton therapy (n = 95). The authors observed that increases in heart V30Gy were associated with significant risks of cardiac toxicity on univariable and multivariable analysis. Additionally, median heart V30Gy values were 57.6%, 26.2%, and 20.0% for 3D-CRT, IMRT, and protons, respectively (p < 0.0001). As shown in Figure 2B, we observed a mean heart V30Gy of 35.6% 95% CI [32.5, 38.7] versus 24.7% 95% CI [23.2, 26.2] for 3D-CRT and IMRT, respectively, further validating their findings and supporting the use of IMRT over 3D-CRT in this setting. The benefits of proton therapy in the setting of esophageal cancer continues to be a topic of significant study. In 2020, a phase IIB randomized trial by Lin et al. conduced at the MD Anderson Cancer Center was published, which randomized 145 patients with locally advanced esophageal cancer to IMRT and proton beam therapy.20 That study utilized a composite toxicity endpoint, where a score was derived from 11 distinct adverse events. The investigators observed that IMRT was associated with a higher toxicity burden compared to proton therapy. However, that trial had several limitations, such as the use of a toxicity endpoint that was not previously validated and a high rate of unevaluable patients in the proton arm, which was likely due to insurance reasons.21 The ongoing NRG GI-006 trial (NCT03801876) is further evaluating this hypothesis with co-primary endpoints of OS and incidence of severe cardiopulmonary adverse events.22 This trial is expected to be completed in 2031.
In 2003, Lee et al. published a retrospective study of 117 patients who underwent preoperative chemoradiotherapy for esophageal cancer at the MD Anderson Cancer Center.8 They observed a higher rate of pulmonary complications for a total lung V10Gy ≥40% compared to < 40% (35% versus 8%; p = 0.014) and for a total lung V15Gy ≥30% compared to < 30% (33% versus 10%; p = 0.036). A similar study performed at the Mayo Clinic found that increasing total lung V10Gy and V15Gy was associated with higher rates of pulmonary complications.23 However, in addition to being retrospective in nature, these studies exclusively utilized 3D-CRT. In our study, where the majority of patients were treated with IMRT, we observed similar findings with respect to the total lung V10 Gy. First, on univariable analysis, exceeding both the protocol specified (V10 Gy ≤40%) and the calculated total lung dose constraints (V10 Gy ≤42%) were associated with worse OS (HR: 1.55; 95% CI: 1.04-2.30; p = 0.03 and HR: 1.88; 95% CI: 1.25-2.83; p = 0.002, respectively). Additionally, on multivariable analysis, a total lung V10 Gy > 42% was associated with worse OS (HR: 1.71; 95% CI: 1.102.66; p = 0.017). These findings are presented in Table 4. While the use of IMRT over 3D-CRT did not confer an increased ability to meet the calculated dose constraint of V10 Gy ≤42%, maintaining the total lung dose within this constraint was associated with an OS of 55.7 months versus 21.7 months (p = 0.004), as shown in Table 3 and Figure 1.
While the recent presentation of the ESOPEC trial (NCT02509286) comparing perioperative fluorouracil, leucovorin, oxaliplatin, and docetaxel (FLOT) chemotherapy versus preoperative chemoradiotherapy per the CROSS regimen demonstrated an OS benefit for the use of perioperative FLOT, there are still many esophageal cancer patients who may not be a candidate for the aggressive FLOT regimen24. In those cases, the CALGB 80803 approach is the “best of both worlds,” combining more effective induction chemotherapy, followed by response-adapted chemoradiation. Thus, there are still patients for whom the dosimetric data for radiation therapy planning will be critical. Our findings demonstrate that IMRT over 3D-CRT is able to lower doses to the heart and lungs, thus minimizing the impact of radiation treatment on OS.
This study does have multiple limitations, and these results should therefore be interpreted in context. First, while we were able to determine new thresholds for each of the volumetric total lung and heart variables that were defined in the study protocol, an analysis of other volumetric thresholds was not performed. Thus, it is possible that the optimal doses received by the lungs and heart to maximize survival in this setting were not captured in this analysis. Second, long-term follow-up and assessment of the risk of late cardiac or pulmonary toxicity was not conducted, limiting the ability to evaluate the long-term impact of cardiac and pulmonary irradiation. Third, the primary study question was not related to radiation modality or dose-volume parameters. Therefore, it is likely that measured and unmeasured covariates were not fully balanced in the radiation modality of dose-volume comparative subgroups. Fourth, a detailed analysis of specific cardiopulmonary toxicities beyond survival was not possible due to the lack of granularity of the adverse event data. Future studies should incorporate these analyses, as well as the impact of radiation doses to cardiac substructures, which have been correlated with cardiopulmonary morbidity. Fifth, while we did not observe any differences in baseline characteristics between the 3D-CRT and IMRT groups, it is possible that interactions between these variables impacted compliance with protocol-specified dose constraints and radiation modality selection. Finally, cause of death was not collected in all cases, so the association of heart and lung doses to mortality from cardiac events cannot be definitively determined.
CONCLUSION
The majority of patients on CALGB 80803 (Alliance) were treated with IMRT plans that met the protocol-specified dose constraints for total lung and heart. Minimizing volume of low dose to the lungs (V10Gy ≤42%) and minimizing dose to the heart (V30Gy ≤21%) were both associated with significant OS benefits. Additionally, utilization of IMRT over 3D-CRT allowed for greater heart sparing. Obtaining a total lung V10Gy ≤42% and heart V30Gy ≤21% is more likely to be achieved with IMRT over 3D-CRT.
Supplementary Material
ACKNOWLEDGEMENTS:
https://acknowledgments.alliancefound.org
FUNDING SOURCES:
Research reported in this publication was supported by the National Cancer Institute of the National Institutes of Health under Award Numbers U10CA180821 and U10CA180882 (to the Alliance for Clinical Trials in Oncology), UG1CA232760, UG1CA233290, and UG1CA233331. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health.
Footnotes
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
CONFLICTS OF INTEREST:
Karyn A. Goodman, MD, MS: Advisory Board, RenovoRx
CLINICALTRIALS.GOV IDENTIFIER: NCT01333033
Data Sharing Statement:
Data are available through the Alliance for Clinical Trials in Oncology
REFERENCES
- 1.Then EO, Lopez M, Saleem S, et al. Esophageal Cancer: An Updated Surveillance Epidemiology and End Results Database Analysis. World J Oncol. Apr 2020;11(2):55–64. doi: 10.14740/wjon1254 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.van Hagen P, Hulshof MC, van Lanschot JJ, et al. Preoperative chemoradiotherapy for esophageal or junctional cancer. N Engl J Med. May 31 2012;366(22):2074–84. doi: 10.1056/NEJMoa1112088 [DOI] [PubMed] [Google Scholar]
- 3.Shapiro J, van Lanschot JJB, Hulshof 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. Sep 2015;16(9):1090–1098. doi: 10.1016/S1470-2045(15)00040-6 [DOI] [PubMed] [Google Scholar]
- 4.Eyck BM, van Lanschot JJB, Hulshof M, et al. Ten-Year Outcome of Neoadjuvant Chemoradiotherapy Plus Surgery for Esophageal Cancer: The Randomized Controlled CROSS Trial. J Clin Oncol. Jun 20 2021;39(18):1995–2004. doi: 10.1200/JCO.20.03614 [DOI] [PubMed] [Google Scholar]
- 5.Siegel RL, Miller KD, Wagle NS, Jemal A. Cancer statistics, 2023. CA Cancer J Clin. Jan 2023;73(1):17–48. doi: 10.3322/caac.21763 [DOI] [PubMed] [Google Scholar]
- 6.Siegel RL, Kratzer TB, Giaquinto AN, Sung H, Jemal A. Cancer statistics, 2025. CA Cancer J Clin. Jan-Feb 2025;75(1):10–45. doi: 10.3322/caac.21871 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Arnold M, Laversanne M, Brown LM, Devesa SS, Bray F. Predicting the Future Burden of Esophageal Cancer by Histological Subtype: International Trends in Incidence up to 2030. Am J Gastroenterol. Aug 2017;112(8):1247–1255. doi: 10.1038/ajg.2017.155 [DOI] [PubMed] [Google Scholar]
- 8.Lee HK, Vaporciyan AA, Cox JD, et al. Postoperative pulmonary complications after preoperative chemoradiation for esophageal carcinoma: correlation with pulmonary dose-volume histogram parameters. Int J Radiat Oncol Biol Phys. Dec 1 2003;57(5):1317–22. doi: 10.1016/s0360-3016(03)01373-7 [DOI] [PubMed] [Google Scholar]
- 9.Wang SL, Liao Z, Vaporciyan AA, et al. Investigation of clinical and dosimetric factors associated with postoperative pulmonary complications in esophageal cancer patients treated with concurrent chemoradiotherapy followed by surgery. Int J Radiat Oncol Biol Phys. Mar 1 2006;64(3):692–9. doi: 10.1016/j.ijrobp.2005.08.002 [DOI] [PubMed] [Google Scholar]
- 10.Garant A, Spears G, Routman D, et al. A Multi-Institutional Analysis of Radiation Dosimetric Predictors of Toxicity After Trimodality Therapy for Esophageal Cancer. Practical Radiation Oncology Jul-Aug 2021;11(4):e415–e425. doi: 10.1016/j.prro.2021.01.004 [DOI] [PubMed] [Google Scholar]
- 11.Tucker SL, Liu HH, Wang S, et al. Dose-volume modeling of the risk of postoperative pulmonary complications among esophageal cancer patients treated with concurrent chemoradiotherapy followed by surgery. Int J Radiat Oncol Biol Phys. Nov 1 2006;66(3):754–61. doi: 10.1016/j.ijrobp.2006.06.002 [DOI] [PubMed] [Google Scholar]
- 12.Boyle J, Ackerson B, Gu L, Kelsey CR. Dosimetric advantages of intensity modulated radiation therapy in locally advanced lung cancer. Adv Radiat Oncol Jan-Mar 2017;2(1):6–11. doi: 10.1016/j.adro.2016.12.006 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Lin SH, Wang L, Myles B, et al. Propensity score-based comparison of long-term outcomes with 3-dimensional conformal radiotherapy vs intensity-modulated radiotherapy for esophageal cancer. Int J Radiat Oncol Biol Phys. Dec 1 2012;84(5):1078–85. doi: 10.1016/j.ijrobp.2012.02.015 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Goodman KA, Ou FS, Hall NC, et al. Randomized Phase II Study of PET Response-Adapted Combined Modality Therapy for Esophageal Cancer: Mature Results of the CALGB 80803 (Alliance) Trial. J Clin Oncol. Sep 1 2021;39(25):2803–2815. doi: 10.1200/JCO.20.03611 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Gagliardi G, Constine LS, Moiseenko V, et al. Radiation dose-volume effects in the heart. Int J Radiat Oncol Biol Phys. Mar 1 2010;76(3 Suppl):S77–85. doi: 10.1016/j.ijrobp.2009.04.093 [DOI] [PubMed] [Google Scholar]
- 16.Kavanagh BD, Pan CC, Dawson LA, et al. Radiation dose-volume effects in the stomach and small bowel. Int J Radiat Oncol Biol Phys. Mar 1 2010;76(3 Suppl):S101–7. doi: 10.1016/j.ijrobp.2009.05.071 [DOI] [PubMed] [Google Scholar]
- 17.Marks LB, Yorke ED, Jackson A, et al. Use of normal tissue complication probability models in the clinic. Int J Radiat Oncol Biol Phys. Mar 1 2010;76(3 Suppl):S10–9. doi: 10.1016/j.ijrobp.2009.07.1754 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Kassambara A, Kosinski M, Biecek P.). survminer: Drawing Survival Curves using ‘ggplot2’. R package version 0.4.9. https://CRAN.R-project.org/package=survminer. 2021; [Google Scholar]
- 19.Hothorn T, Lausen B. On the Exact Distribution of Maximally Selected Rank Statistics. Science Direct Working Paper No S1574-0358(04)70152-5. 2002; [Google Scholar]
- 20.Lin SH, Hobbs BP, Verma V, et al. Randomized Phase IIB Trial of Proton Beam Therapy Versus Intensity-Modulated Radiation Therapy for Locally Advanced Esophageal Cancer. J Clin Oncol. May 10 2020;38(14):1569–1579. doi: 10.1200/JCO.19.02503 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Simone CB 2nd. First Randomized Trial Supporting the Use of Proton Over Photon Chemoradiotherapy in Esophageal Cancer. J Clin Oncol. Sep 10 2020;38(26):2952–2955. doi: 10.1200/JCO.20.01405 [DOI] [PubMed] [Google Scholar]
- 22.Comparing Proton Therapy to Photon Radiation Therapy for Esophageal Cancer. https://clinicaltrials.gov/study/NCT03801876. [Google Scholar]
- 23.Schallenkamp JM, Miller RC, Brinkmann DH, Foote T, Garces YI. Incidence of radiation pneumonitis after thoracic irradiation: Dose-volume correlates. Int J Radiat Oncol Biol Phys. Feb 1 2007;67(2):410–6. doi: 10.1016/j.ijrobp.2006.09.030 [DOI] [PubMed] [Google Scholar]
- 24.Hoeppner J, Bruner T, Lordick F, et al. Prospective randomized multicenter phase III trial comparing perioperative chemotherapy (FLOT protocol) to neoadjuvant chemoradiation (CROSS protocol) in patients with adenocarcinoma of the esophagus (ESOPEC trial). presented at: American Society for Clinical Oncology 2024 Meeting; June 5, 2024. 2024; Chicago, IL, USA. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data Availability Statement
Data are available through the Alliance for Clinical Trials in Oncology