- Email: [email protected]
A Comparison of Patient-Reported Health-Related Quality of Life During Proton Versus Photon Chemoradiation Therapy for Esophageal Cancer
Practical Radiation Oncology (2019) xx, 1-8
www.practicalradonc.org
Basic Original Report
A Comparison of Patient-Reported Health-Related Quality of Life During Proton Versus Photon Chemoradiation Therapy for Esophageal Cancer Aurelie Garant MD a, Thomas J. Whitaker PhD a, Grant M. Spears BS b, David M. Routman MD a, William S. Harmsen MS b, Tyler J. Wilhite MD a, Jonathan B. Ashman MD, PhD c, Terence T. Sio MD c, William G. Rule MD c, Michelle A. Neben Wittich MD a, James A. Martenson MD a, Erik J. Tryggestad MD a, Harry H. Yoon MD d, Shanda Blackmon MD e, Kenneth W. Merrell MD a, Michael G. Haddock MD a, Christopher L. Hallemeier MD a,* Departments of aRadiation Oncology and bBiomedical Statistics and Informatics, Mayo Clinic, Rochester, Minnesota; c Department of Radiation Oncology, Mayo Clinic, Phoenix, Arizona; and Departments of dMedical Oncology and e Surgery, Mayo Clinic, Rochester, Minnesota Received 22 March 2019; revised 1 July 2019; accepted 2 July 2019
Abstract Purpose: The purpose of this study was to compare Functional Assessment of Cancer TherapyEsophagus (FACT-E) questionnaire changes during proton (PRT) or photon (XRT) chemoradiation therapy (CRT) for esophageal cancer (EC). Methods and Materials: We reviewed patients enrolled in a prospective registry who received preoperative or definitive CRT for EC. Patients completed the FACT-E before CRT and during the last week of CRT. Analysis of variance testing was used to assess associations between patient and treatment characteristics and FACT-E score changes. Results: One hundred twenty-five patients completed a baseline and posttreatment FACT-E; 63 received XRT and 62 received PRT. The mean age was 65 years; the PRT group was older (68 vs 64 years, P Z .0063). The following characteristics were similar between cohorts: 83% male, 78% adenocarcinoma, and 89% stage II-III. The radiation therapy prescription dose was higher in the PRT group (50 Gy in 94% vs 67%, P < .0001), whereas the median clinical target volume was smaller in the PRT group (553 vs 668 cm3, P Z .013). Most (96%) received concurrent weekly carboplatin-paclitaxel. The mean FACT-E score was 136.3 (standard deviation [SD] Sources of support: This work had no specific funding. Disclosures: Dr Blackmon reports nonfinancial support from Boston Scientific, Ethicon, Medtronic, and Olympus outside the submitted work. All other authors have no disclosures to declare. * Corresponding author. E-mail address: [email protected] (C.L. Hallemeier). https://doi.org/10.1016/j.prro.2019.07.003 1879-8500/Ó 2019 American Society for Radiation Oncology. Published by Elsevier Inc. All rights reserved.
2
Practical Radiation Oncology: --- 2019
A. Garant et al
21.0) at baseline and 119.6 (SD 24.8) post-CRT, with mean change of e16.7 (SD 19.8). Baseline scores were comparable between XRT and PRT groups (135.9 vs 136.7, P Z .82). On univariate and multivariate analyses, less mean decline in FACT-E score was observed for PRT versus XRT (e12.7 vs e20.6, P Z .026) and for trimodality versus definitive therapy (e13.0 vs e22.5, P Z .008). Conclusions: For patients receiving CRT for EC, PRT was associated with less decline in FACT-E scores compared with XRT. Ó 2019 American Society for Radiation Oncology. Published by Elsevier Inc. All rights reserved.
Introduction In the era of optimized staging technologies and patient selection, curative therapies for locally advanced esophageal cancer lead to up to 50% survival at 5 years in contemporary randomized clinical trials.1 Current curative approaches in the management of esophageal cancer involve multimodality therapy. Indeed, either preoperative chemoradiation therapy (CRT) followed by oncologic resection or definitive CRT are acceptable standards and may be adapted to each patient’s health profile.1,2 Photon radiation therapy (XRT) using either 3-dimensional conformal or intensity modulated radiation therapy (IMRT) is the most commonly used RT modality for esophageal cancer. However, there is increasing interest in the potential use of proton RT (PRT) for esophageal cancer owing to favorable dose distribution with reduction in integral dose delivered to the total body, heart, lungs, stomach, and bowel.3-6 Preliminary retrospective data suggest that PRT is a feasible and safe technique for esophageal cancer, with suggestion of reduced treatmentrelated adverse effects and favorable survival outcomes.7,8 However, there are no studies comparing patient-reported health-related quality of life (HRQoL) for patients with esophageal cancer treated with PRT versus XRT. The purpose of this analysis was to compare HRQoL during CRT for patients with esophageal cancer receiving PRT or XRT. HRQoL was assessed using the Functional Assessment of Cancer Therapy-Esophageal (FACT-E) questionnaire, a validated instrument consisting of a general questionnaire (FACT-G) and an esophageal cancer subscale (ECS).9,10 Owing to the more favorable dose distribution observed with PRT versus XRT for esophageal cancer, we hypothesized that PRT would be associated with less decline in HRQoL during CRT compared with XRT.
Methods and Materials Patient selection Starting in June 2015, patients treated with curativeintent RT at 2 campuses from 1 institution were approached by a research coordinator at the time of
simulation or first RT treatment for potential participation in a patient outcomes registry. For this institutional review boardeapproved prospective study, patients provided written consent for use of their data for research purposes. Patients were administered HRQoL studies on a portable electronic device at registration (simulation or first RT treatment) and during the last week of RT. For this analysis, inclusion criteria were esophageal or gastroesophageal junction carcinoma treated with curative-intent preoperative or definitive RT between June 2015 and February 2018. Exclusion criteria were prior esophagectomy or gastrectomy, prior thoracic RT, a palliative course of treatment (planned RT dose <40 Gy), receipt of both PRT and XRT, lack of concomitant chemotherapy, or no pretreatment questionnaire.
Treatment Patients received CRT in the preoperative or definitive setting. Pretreatment feeding tubes were uncommonly used (<10% of patients), and typically only for patients with complete esophageal obstruction. CRT was categorized as preoperative or definitive based on actual treatment received (ie, whether the patient actually underwent esophagectomy). Standard chemotherapy consisted of weekly intravenous carboplatin 50 mg/m2 and paclitaxel titrated to an area under the curve of 2. Decision on whether to use PRT or XRT was based primarily on insurance coverage and treatment scheduling availability. All patients underwent simulation with a 4-dimensional CT scan. For both modalities, the clinical tumor volume (CTV) included the primary tumor with a 3- to 4-cm mucosal expansion and a 1-cm radial expansion as well as involved regional lymph nodes with a 1-cm expansion. The CTV typically included the celiac lymph nodes for distal esophagus or gastroesophageal junction tumors and the supraclavicular lymph nodes for upper esophagus tumors. CTVs were typically modified to exclude uninvolved adjacent organs at risk, including the heart, lungs, bones, and bowel. For XRT, the planning target volume was a 5-mm uniform expansion of the CTV. For PRT, robust treatment planning was used with a goal to have at least 95% of the CTVs receiving 95% of the prescription dose with translational shifts of 5 mm in x, y, and z directions and range uncertainty of 3%. PRT was
Practical Radiation Oncology: --- 2019
Figure 1
Photon or proton RT for esophageal cancer
3
Example of photon (left) and proton (right) dose distributions for a sample patient.
administered using pencil beam spot scanning, typically with 2 posterior oblique fields. XRT typically consisted of volumetric modulated arc RT (70% of patients) or 3dimensional conformal RT (30% of patients). Examples of dosimetric features for PRT and XRT are depicted in Figure 1. For both PRT and XRT, daily image guidance consisted of orthogonal kilovoltage x-rays. The most common RT dose and fractionation regimens used were 50 Gy in 25 fractions or 50.4 Gy in 28 fractions. For patients receiving preoperative RT, 41.4 Gy in 23 fractions was also an option, which was more commonly used with XRT, especially if dose constraints for the heart (mean 26 Gy) or lung (mean 15Gy) could not be met with a dose of 50 Gy.
HRQoL questionnaires At baseline and at the end of CRT treatment, patients were asked to complete the FACT-E survey. The FACT-E contains the FACT-G and the ECS, which have been validated for evaluation of patient-reported outcomes in the context of esophageal cancer treatments.9,10 The FACT-G covers domains of physical well-being (PWB), social and family well-being (SWB), emotional wellbeing (EWB), and functional well-being (FWB) for adults with cancer.9,11 The ECS contains 17 items addressing symptoms and problems related to esophageal cancer. The FACT-E trial outcome index includes the PWB, FWB, and ECS subscales. The FACT-E takes approximately 10 minutes to complete. The FACT-E total score is the sum of the FACT-G scores and the ECS; higher scores indicate a better health-related QoL. The maximum FACT-E score is 176 points.
Statistics Baseline patient and treatment characteristics were compared between the XRT and PRT cohorts using the 2-sample t test or Fisher’s exact test. Change in HRQoL scores from baseline to end of treatment were assessed using the 2-sample t test. Univariable and multivariable linear regression analyses were used to assess for associations between patient and treatment characteristics and HRQoL score changes. For all statistical tests, a P value of <.05 was considered statistically significant. We performed sensitivity analyses, including patients who had completed the baseline questionnaire but not the end-of-CRT questionnaire. The patient’s chart was reviewed, and the end-of-treatment questionnaire was determined to be “missing at random” or “missing not at random.” Patients were determined to be “missing not at random” if documentation suggested clinical deterioration with acute grade 3 toxicity based on the Common Terminology Criteria for Adverse Events (v4.03), hospital admission, or a non-cancer-related event precluding patient participation. Patients were determined to be “missing at random” if the previous criteria were not met and documentation indicated that the patient would have been capable of filling out a questionnaire at the end of treatment. Two imputation models were designed to take missing posttreatment questionnaires into account. In both models, the patients with “missing at random” questionnaires were assigned a predicted score using a regression model with age, sex, and baseline score as covariates. In the first model, the patients with “missing not at random” questionnaires were attributed a score of the worst recorded score among patients with posttreatment
4
Practical Radiation Oncology: --- 2019
A. Garant et al Table 1
Patient and treatment characteristics for patients with a baseline and end-of-treatment FACT-E questionnaire
Age at diagnosis, y, mean (SD) Sex, no. (%) Female Male Tumor location, no. (%) Upper or middle Lower or gastroesophageal junction Histology, no. (%) Adenocarcinoma Squamous Other* Stage, no. (%) Iy II III IVz Not applicable* Therapy received, no. (%) Definitive Trimodality Concurrent chemotherapy, no. (%) Carboplatin/Paclitaxel Otherx Clinical target volume, cm3, mean (SD) Radiation therapy fractions, mean (SD) Radiation therapy dose received, no. (%) <5000 cGy 4140 cGy 5000 cGy 5000 cGy 5040 cGy
Total (n Z 125)
Proton (n Z 62)
Photon (n Z 63)
65.9 (10.6)
68.9 (9.9)
62.9 (10.5)
21 (16.8) 104 (83.2)
9 (14.5) 53 (85.5)
P Value .0014 .63
12 (19.0) 51 (81.0) 1.0
107 (85.6) 18 (15.4)
53 (85.5) 9 (14.5)
54 (85.7) 9 (14.3) .30
97 (77.6) 17 (13.6) 11 (8.8)
51 (82.3) 8 (12.9) 3 (4.8)
46 (73.0) 9 (14.3) 8 (12.7)
9 21 88 4
5 13 42 2
4 8 46 2
.66 (7.4) (17.2) (72.1) (3.3) 3
(8.1) (21.0) (67.7) (3.2) 0
(6.7) (13.3) (76.7) (3.3) 3 .14
49 (39.2) 76 (60.8)
20 (32.3) 42 (67.7)
29 (46.0) 34 (54.0)
120 (96.0) 5 (4.0) 611 (258) 25.2 (1.6)
61 (98.4) 1 (1.6) 553 (238) 24.9 (0.6)
59 (93.6) 4 (6.4) 668 (267) 25.4 (2.2)
25 (20.0)
4 (6.5) 2 58 (93.5) 58 1
21 (33.3) 19 42 (66.7) 20 20
.21
100 (80.0)
.013 .14 .0002
Abbreviations: FACT-E Z Functional Assessment of Cancer Therapy-Esophagus; SD Z standard deviation. * Localized neuroendocrine or small cell carcinoma. y Patients with stage I disease were medically inoperable or declined esophagectomy. z Patients with stage IV disease had limited metastatic spread to contiguous lymph nodes or organs and were treated with curative-intent radiation therapy and chemotherapy. x Three patients received cisplatin and etoposide, and 2 patients received fluorouracil and oxaliplatin.
data minus 1 point; in the second model, the patients with “missing not at random” questionnaires were attributed a score of 2 out of 4 points (50%) for each question.
Results Between June 2015 and February 2018, 189 patients met the study inclusion criteria and filled out a baseline questionnaire. In total, 125 patients completed both a baseline and end-of-treatment FACT-E. All patients received the planned RT dose except for 1 XRT patient who stopped after 34.2 Gy. The end-of-treatment survey was completed a median of 1 day (interquartile range [IQR], 0-3) before the last day of RT, with no significant difference observed between the PRT (median 1 day; IQR, 0-3) and XRT (median 1 day; IQR, 0-2) cohorts
(P Z .15 by KruskaleWallis test). Patient and treatment characteristics are summarized in Table 1. There were significant differences observed between PRT and XRT cohorts in regard to age, CTV, and prescription dose. Specifically, the PRT cohort (compared with the XRT cohort) had an older mean age, a smaller mean CTV, and a higher prescription dose. The mean heart dose was 7.0 Gy for the PRT cohort and 20.2 Gy for the XRT cohort (P < .0001). The mean lung dose was 3.6 Gy for the PRT cohort and 8.9 Gy for the XRT cohort (P < .0001). Results of the FACT-E surveys are summarized in Table 2. The mean FACT-E score was 136.3 points (standard deviation [SD] 21.0) at baseline. The baseline mean FACT-E score was similar for patients receiving preoperative versus definitive therapy (137.2 vs 135.0, P Z .57) and for patients receiving PRT versus XRT
Practical Radiation Oncology: --- 2019 Table 2
Photon or proton RT for esophageal cancer
5
FACT-E results for patients with baseline and end-of-treatment questionnaires (mean, standard deviation)
FACT-E total scores Baseline Post-treatment FACT-G total scores Baseline Post-treatment Differences between baseline and post-treatment scores FACT-E FACT-G FACT-E Trial Outcome Index Esophageal cancer subscale Emotional well-being subscale Functional well-being subscale Physical well-being subscale Social and family well-being subscale
Total (n Z 125)
Proton (n Z 62)
Photon (n Z 63)
P Value
136.3 (21.0) 119.6 (24.8)
136.7 (23.1) 124.0 (26.9)
135.9 (18.8) 115.3 (21.9)
.82 .048
87.0 (13.7) 77.4 (16.3)
87.2 (14.5) 80.3 (16.5)
86.9 (13.1) 74.5 (15.8)
.89 .047
e16.7 e9.7 e16.1 e7.0 e0.4 e3.1 6.0 e0.1
(19.8) (12.7) (17.2) (11.1) (4.4) (5.6) (5.4) (4.3)
e12.7 e6.9 e13.1 e5.8 0.0 e1.6 5.6 0.4
(19.9) (12.5) (17.4) (10.5) (4.4) (6.1) (5.1) (4.1)
e20.6 e12.4 e19.2 e8.3 e0.8 e4.6 e6.3 e0.6
(19.1) (12.3) (16.6) (11.7) (4.3) (4.7) (5.6) (4.3)
.026 .015 .048 .22 .29 .0032 .43 .20
Abbreviations: FACT-E Z Functional Assessment of Cancer Therapy-Esophagus; FACT-G Z Functional Assessment of Cancer Therapy-General.
(136.7 vs 135.9, P Z .82). The mean FACT-E score for the whole cohort was 119.6 points (SD 24.8) at the end of CRT, with mean change of e16.7 (SD 19.8). The greatest numerical declines were noted in the ECS (mean change of e7), PWB (mean change of e6), and FWB (mean change of e3.1), whereas there was no significant change in social and emotional well-being. In the univariable linear regression model (Table 3), less mean decline in FACT-E was observed for PRT versus XRT (-12.7 vs -20.6, P Z .026) and for trimodality therapy versus definitive therapy (-13.0 vs -22.5, P Z .008). There was also significantly less mean decline with PRT versus XRT for the FACT-G (-6.9 vs -12.4, P Z .015) and the FACT-E Trial Outcome Index (-13.1 vs -19.2, P Z .048). For all survey subscales, the median decline was numerically less in the PRT versus XRT cohort, although this was only statistically different for the FWB Subscale (-1.6 vs -4.6, P Z .0032). In the multivariable linear regression model, PRT (vs XRT, P Z .026) and trimodality therapy (vs definitive therapy, P Z .004) were associated with less decline in FACT-E. Sensitivity analyses were performed on the entire cohort of 189 patients with a baseline FACT-E survey, including the 64 patients (34%) who did not complete a posttreatment survey (patient and treatment characteristics are outlined in Table E1, available online at https://doi. org/10.1016/j.prro.2019.07.003). This included 20 patients who received PRT and 44 patients who received XRT. Among the 64 missing posttreatment surveys, 48 surveys were determined to be missing at random and 16 were missing not at random. The percentage of surveys missing at random versus not at random was not different for the PRT versus XRT cohorts. When imputing data for missing questionnaires under several scenarios, the
different declines in scores of PRT and XRT cohorts remained significant. In the scenario where all nonrandomly missing questionnaires were imputed with a posttreatment score of the worst recorded score in the database minus 1, the XRT versus PRT mean score declines were -23.3 vs -16.0, respectively (P Z .022). In the scenario where all nonrandomly missing questionnaires were imputed with a posttreatment score of 50%, the XRT versus PRT median score declines were -20.7 and -13.9, respectively (P Z .015). These changes in HRQoL scores are depicted in Table 4 and Figure 2.
Discussion In this analysis of patients with esophageal cancer receiving curative intent neoadjuvant or definitive CRT, we observed that HRQoL declined significantly from pretreatment to the end of CRT, consistent with previous observations.12 Importantly, RT modality was associated with decline in HRQoL, with significantly less decline in patients who received PRT versus XRT. This is one of the largest studies to date assessing patient-reported HRQoL during CRT for esophageal cancer. To our knowledge, this is the only study thus far assessing for association between RT modality and HRQoL during CRT for esophageal cancer. Additionally, this is one of few studies comparing HRQoL with XRT versus PRT for any type of cancer.13 Strengths of this study include the relatively large cohort of patients with esophageal cancer treated with CRT over a relatively short time interval; thus, treatment regimens were relatively homogenous, minimizing risk of bias. A vast majority of patients with esophageal cancer treated at our
6
Practical Radiation Oncology: --- 2019
A. Garant et al Table 3
Univariable and multivariable analyses of variables potentially associated with FACT-E change
Variable
Univariable
Age, y Male (vs female) Adenocarcinoma (vs squamous cell) stage III or IV (vs I or II) Trimodality (vs definitive) CTV (per 100 cm3) RT dose 5000 cGy (vs < 5000 cGy) Proton (vs photon)
Multivariable
Estimate, points
Standard error, points
P Value
Estimate, points
Standard error, points
P Value
0.12 3.22 1.94
0.17 4.76 5.13
.49 .50 .71
0.14 1.49 0.19
0.18 5.62 5.3
.43 .79 .97
5.69 9.55 0.090 4.27
4.17 3.55 0.69 4.44
.18 .008 .90 .34
4.40 11.07 1.22 0.47
4.41 3.74 0.83 4.78
.32 .004 .14 .92
7.89
3.49
.026
9.21
4.07
.026
Abbreviations: CTV Z clinical target volume; FACT-E Z Functional Assessment of Cancer Therapy-Esophagus; RT Z radiation therapy.
institutions (95%) had competed baseline HRQoL, minimizing risk of selection bias. To our knowledge, there have not been any studies that have defined minimally important difference or clinically meaningful change in the FACT-E in patients with esophageal cancer. In a study of patients with gastric cancer, a mean change in FACT-G score of 3.9, 7.8, and 11.8 points represented a small, medium, or large change in QoL, respectively.14 A mean change in the FACT-Ga (including the FACT-G and gastric cancer subscale) score of 7.5, 15.1, and 22.6 points represented a small, medium, or large change in QoL, respectively. Other studies have defined a one-third change in standard deviation in FACT scores as clinically meaningful change.15,16 Using data derived from the present study, a one-third standard deviation would correspond to approximately 5 points for the FACT-G and approximately 7.5 points for the FACT-E. Therefore, the observed statistically significant differences of 5.5 points for the FACT-G and 7.9 points for the FACT-E between XRT and PRT patients in our study likely represents a small but clinically meaningful change in QOL. There are limitations to our analysis. Given that this study was retrospective, there are potential selection biases and confounders that are difficult to account for in a nonrandomized setting. Specifically, there were significant differences between the PRT and XRT cohorts with
Table 4
respect to patient age, RT prescription dose, and CTV. The discrepancy in age between the PRT and XRT cohorts relates to the fact that during the study period, most private insurance carriers did not cover PRT for esophageal cancer, whereas our Medicare carrier did. However, patient age, RT prescription dose, and CTV were not associated with a change in FACT-E score during CRT, suggesting that imbalances in the covariates between the PRT and XRT cohorts were unlikely to explain the observed change in FACT-E. Similarly, a previous study did not demonstrate an association between patient age and change in QOL during CRT for esophageal cancer.17 Furthermore, the baseline FACT-E scores were similar for the PRT and XRT cohorts (136.7 vs 135.9, P Z .82), arguing against clinically significant imbalance in the pretreatment patient or disease characteristics. Sixty-four patients (34%) were excluded from the primary analysis because they did not complete an end-of-treatment HRQoL questionnaire. Unfortunately, missing data are a common occurrence in QoL research in patients with cancer.18-20 The rate of missing data at the end of treatment in our study is similar to that observed in recent prospective studies of QoL during CRT for esophageal cancer, in which end-of-treatment surveys were missing in 18% to 41% of patients.21-23 We performed sensitivity analyses to account for missing end-of-treatment surveys using common imputation methods, which also showed a
FACT-E results including patients with imputed scores for missing questionnaires, mean (standard deviation)
Differences between baseline and post-treatment scores
Total (n Z 189)
Proton (n Z 82)
Photon (n Z 107)
P Value
FACT-E, imputation method 1 FACT-E, imputation method 2 FACT-G, imputation method 1 FACT-G, imputation method 2
20.1 17.8 13.4 10.5
16.0 13.9 10.4 7.9
23.3 20.7 15.7 12.5
.022 .015 .036 .011
(21.8) (19.0) (17.2) (12.3)
(22.4) (20.7) (16.3) (12.7)
(20.9) (17.1) (17.6) (11.8)
Abbreviations: FACT-E Z Functional Assessment of Cancer Therapy-Esophagus; FACT-G Z Functional Assessment of Cancer Therapy-General.
Practical Radiation Oncology: --- 2019
Photon or proton RT for esophageal cancer
status.26-28 Djarv et al studied QoL changes in 132 patients with esophageal or gastric cancer who had received curative-intent CRT or surgery.29 QoL was assessed using the European Organization for Research and Treatment of Cancer QLQ-C30 at baseline and after 6 months. In a Cox proportional hazard analyses adjusted for age, sex, performance status, tumor stage, histology, disease site, baseline HRQoL, and treatment, better recovery of physical function at 6 months was associated with better survival; a 10-point increase in problems with pain and fatigue was associated with a higher risk of death. Further evaluation is needed to better establish the associations between HRQoL and oncologic outcomes.
175
150
FACT E Score
125
100
75
50
25
0
7
Radiation type
Conclusions
Photon Proton Baseline
Post−Tx
Patients with both surveys
Baseline
Post−Tx
All Patients, Imputation method 1
Baseline
Post−Tx
All Patients, Imputation method 2
Figure 2 FACT-E score changes for proton versus photon cohorts (dot Z mean; bracket Z standard deviation).
significant difference in FACT-E scores for the PRT versus XRT cohorts. We hypothesize that the observation of less decline in HRQoL with PRT versus XRT is related to the more favorable dose distribution of PRT versus XRT for esophageal cancer, as illustrated in Figure 1. Specifically, PRT (compared with XRT) reduces integral dose to most organs at risk including the heart, lungs, stomach, bowel, liver, and kidneys.3-6 Emerging clinical data suggest PRT may be associated with a lower incidence of treatmentrelated adverse effects including severe lymphopenia and favorable survival outcomes compared with XRT.8,24,25 Reduction in dose to the stomach, bowel, circulating blood cells, and total body would be anticipated to reduce acute treatment-related nausea, anorexia, and fatigue, all components that affect HRQoL. An ongoing phase 2 randomized trial is comparing PRT versus XRT for esophageal cancer (NCT01512589). Endpoints include toxicity burden, HRQoL, and survival. Additionally, a phase 3 randomized controlled trial of PRT versus XRT for esophageal cancer is now open through NRG Oncology (NRG-GI006, NCT03801876). These prospective trials will provide higher-quality data comparing adverse events, HRQoL, and survival with PRT versus XRT. Several investigators have found that HRQoL scores have prognostic implications for patients undergoing therapy for esophageal cancer. In a pooled analysis of 128 patients in 4 prospective, nonrandomized studies of patients with esophageal cancer who received trimodality or definitive therapy, Kidane et al found that pretreatment FACT-E scores had better discrimination for survival than Eastern Cooperative Oncology Group performance
In this observational study of patients with esophageal cancer undergoing curative-intent CRT, use of PRT was associated with significantly less decline in FACT-E scores during treatment compared with XRT. Although the integration HRQoL research is challenging in the outpatient clinic workflow, randomized trials are needed to confirm our findings, and further studies are needed to determine whether changes in HRQoL during CRT are associated with long-term QoL, toxicity, and survival.
Supplementary Data Supplementary material for this article can be found at https://doi.org/10.1016/j.prro.2019.07.003.
References 1. Shapiro J, van Lanschot JJB, Hulshof MCCM, 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. 2. Cooper JS, Guo MD, 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. 3. Chuong MD, Hallemeier CL, Jabbour SK, et al. Improving outcomes for esophageal cancer using proton beam therapy. Int J Radiat Oncol Biol Phys. 2016;95:488-497. 4. Lin SH, Hallemeier CL, Chuong M. Proton beam therapy for the treatment of esophageal cancer. Chin Clin Oncol. 2016;5:53. 5. Lin SH, Komaki R, Liao Z, et al. Proton beam therapy and concurrent chemotherapy for esophageal cancer. Int J Radiat Oncol Biol Phys. 2012;83:e345-351. 6. Xi M, Lin SH. Recent advances in intensity modulated radiotherapy and proton therapy for esophageal cancer. Expert Rev Anticancer Ther. 2017;17:635-646. 7. Lin SH, Merrell KW, Shen J, et al. Multi-institutional analysis of radiation modality use and postoperative outcomes of neoadjuvant chemoradiation for esophageal cancer. Radiother Oncol. 2017;123: 376-381. 8. Xi M, Xu C, Liao Z, et al. Comparative outcomes after definitive chemoradiotherapy using proton beam therapy versus intensity
8
9.
10.
11.
12.
13.
14.
15.
16.
17.
18.
19.
A. Garant et al modulated radiation therapy for esophageal cancer: A retrospective, single-institutional analysis. Int J Radiat Oncol Biol Phys. 2017;99: 667-676. Cella DF, Tulsky DS, Gray G, et al. The Functional Assessment of Cancer Therapy scale: development and validation of the general measure. J Clin Oncol. 1993;11:570-579. Darling G, Eton DT, Sulman J, Casson AG, Celia D. Validation of the functional assessment of cancer therapy esophageal cancer subscale. Cancer. 2006;107:854-863. Overcash J, Extermann M, Parr J, Perry J, Balducci L. Validity and reliability of the FACT-G scale for use in the older person with cancer. Am J Clin Oncol. 2001;24:591-596. Revicki D, Hays RD, Cella D, Sloan J. Recommended methods for determining responsiveness and minimally important differences for patient-reported outcomes. J Clin Epidemiol. 2008;61:102-109. Verma V, Simone CB, Mishra MV. Quality of life and patientreported outcomes following proton radiation therapy: A systematic review. J Natl Cancer Inst. 2017;110:341-353. Garland SN, Pelletier G, Lawe A, et al. Prospective evaluation of the reliability, validity, and minimally important difference of the functional assessment of cancer therapy-gastric (FACT-Ga) qualityof-life instrument. Cancer. 2011;117:1302-1312. Steel JL, Eton DT, Cella D, Olek MC, Carr BI. Clinically meaningful changes in health-related quality of life in patients diagnosed with hepatobiliary carcinoma. Ann Oncol. 2006;17:304-312. Cella D, Eton DT, Fairclough DL, et al. What is a clinically meaningful change on the Functional Assessment of Cancer Therapy-Lung (FACT-L) Questionnaire? Results from Eastern Cooperative Oncology Group (ECOG) Study 5592. J Clin Epidemiol. 2002;55:285-295. Trudel JG, Sulman J, Atenafu EG, Kidane B, Darling GE. Longitudinal evaluation of trial outcome index scores in patients with esophageal cancer. Ann Thorac Surg. 2016;102:269-275. Donaldson GW, Moinpour CM. Learning to live with missing quality-of-life data in advanced-stage disease trials. J Clin Oncol. 2005;23:7380-7384. Macefield RC, Jacobs M, Blencowe NS, et al. The case for a HRQL core outcome set: Outcome reporting bias in oesophageal cancer studies. Trials. 2011;12:A77.
Practical Radiation Oncology: --- 2019 20. Ter Veer E, van Kleef JJ, Sprangers MAG, Haj Mohammad N, van Oijen MGH, van Laarhoven HWM. Reporting of health-related quality of life in randomized controlled trials involving palliative systemic therapy for esophagogastric cancer: A systematic review. Gastric Cancer. 2018;21:1-13. 21. Noordman BJ, Verdam MGE, Lagarde SM, et al. Effect of neoadjuvant chemoradiotherapy on health-related quality of life in esophageal or junctional cancer: Results from the randomized CROSS trial. J Clin Oncol. 2018;36:268-275. 22. Rees J, Hurt CN, Gollins S, et al. Patient-reported outcomes during and after definitive chemoradiotherapy for oesophageal cancer. Br J Cancer. 2015;113:603. 23. Bascoul-Mollevi C, Gourgou S, Galais MP, et al. Health-related quality of life results from the PRODIGE 5/ACCORD 17 randomised trial of FOLFOX versus fluorouracilecisplatin regimen in oesophageal cancer. Eur J Cancer. 2017;84:239-249. 24. Davuluri R, Jiang W, Fang P, et al. Lymphocyte nadir and esophageal cancer survival outcomes after chemoradiation therapy. Int J Radiat Oncol Biol Phys. 2017;99:128-135. 25. Wang J, Wei C, Tucker SL, et al. Predictors of postoperative complications after trimodality therapy for esophageal cancer. Int J Radiat Oncol Biol Phys. 2013;86:885-891. 26. Kidane B, Sulman J, Xu W, et al. Pretreatment quality-of-life score is a better discriminator of oesophageal cancer survival than performance status. Eur J Cardiothorac Surg. 2017;51: 148-154. 27. Kidane B, Sulman J, Xu W, et al. Patient-derived score better discriminator of overall survival than clinician-assigned ECOG status. Interact Cardiovasc Thorac Surg. 2015;21:S33S34. 28. Kidane B, Sulman J, Xu W, et al. Baseline measure of healthrelated quality of life (Functional Assessment of Cancer TherapyEsophagus) is associated with overall survival in patients with esophageal cancer. J Thorac Cardiovasc Surg. 2016;151:15711580. 29. Djarv T, Metcalfe C, Avery KN, Lagergren P, Blazeby JM. Prognostic value of changes in health-related quality of life scores during curative treatment for esophagogastric cancer. J Clin Oncol. 2010; 28:1666-1670.
www.practicalradonc.org
Basic Original Report
A Comparison of Patient-Reported Health-Related Quality of Life During Proton Versus Photon Chemoradiation Therapy for Esophageal Cancer Aurelie Garant MD a, Thomas J. Whitaker PhD a, Grant M. Spears BS b, David M. Routman MD a, William S. Harmsen MS b, Tyler J. Wilhite MD a, Jonathan B. Ashman MD, PhD c, Terence T. Sio MD c, William G. Rule MD c, Michelle A. Neben Wittich MD a, James A. Martenson MD a, Erik J. Tryggestad MD a, Harry H. Yoon MD d, Shanda Blackmon MD e, Kenneth W. Merrell MD a, Michael G. Haddock MD a, Christopher L. Hallemeier MD a,* Departments of aRadiation Oncology and bBiomedical Statistics and Informatics, Mayo Clinic, Rochester, Minnesota; c Department of Radiation Oncology, Mayo Clinic, Phoenix, Arizona; and Departments of dMedical Oncology and e Surgery, Mayo Clinic, Rochester, Minnesota Received 22 March 2019; revised 1 July 2019; accepted 2 July 2019
Abstract Purpose: The purpose of this study was to compare Functional Assessment of Cancer TherapyEsophagus (FACT-E) questionnaire changes during proton (PRT) or photon (XRT) chemoradiation therapy (CRT) for esophageal cancer (EC). Methods and Materials: We reviewed patients enrolled in a prospective registry who received preoperative or definitive CRT for EC. Patients completed the FACT-E before CRT and during the last week of CRT. Analysis of variance testing was used to assess associations between patient and treatment characteristics and FACT-E score changes. Results: One hundred twenty-five patients completed a baseline and posttreatment FACT-E; 63 received XRT and 62 received PRT. The mean age was 65 years; the PRT group was older (68 vs 64 years, P Z .0063). The following characteristics were similar between cohorts: 83% male, 78% adenocarcinoma, and 89% stage II-III. The radiation therapy prescription dose was higher in the PRT group (50 Gy in 94% vs 67%, P < .0001), whereas the median clinical target volume was smaller in the PRT group (553 vs 668 cm3, P Z .013). Most (96%) received concurrent weekly carboplatin-paclitaxel. The mean FACT-E score was 136.3 (standard deviation [SD] Sources of support: This work had no specific funding. Disclosures: Dr Blackmon reports nonfinancial support from Boston Scientific, Ethicon, Medtronic, and Olympus outside the submitted work. All other authors have no disclosures to declare. * Corresponding author. E-mail address: [email protected] (C.L. Hallemeier). https://doi.org/10.1016/j.prro.2019.07.003 1879-8500/Ó 2019 American Society for Radiation Oncology. Published by Elsevier Inc. All rights reserved.
2
Practical Radiation Oncology: --- 2019
A. Garant et al
21.0) at baseline and 119.6 (SD 24.8) post-CRT, with mean change of e16.7 (SD 19.8). Baseline scores were comparable between XRT and PRT groups (135.9 vs 136.7, P Z .82). On univariate and multivariate analyses, less mean decline in FACT-E score was observed for PRT versus XRT (e12.7 vs e20.6, P Z .026) and for trimodality versus definitive therapy (e13.0 vs e22.5, P Z .008). Conclusions: For patients receiving CRT for EC, PRT was associated with less decline in FACT-E scores compared with XRT. Ó 2019 American Society for Radiation Oncology. Published by Elsevier Inc. All rights reserved.
Introduction In the era of optimized staging technologies and patient selection, curative therapies for locally advanced esophageal cancer lead to up to 50% survival at 5 years in contemporary randomized clinical trials.1 Current curative approaches in the management of esophageal cancer involve multimodality therapy. Indeed, either preoperative chemoradiation therapy (CRT) followed by oncologic resection or definitive CRT are acceptable standards and may be adapted to each patient’s health profile.1,2 Photon radiation therapy (XRT) using either 3-dimensional conformal or intensity modulated radiation therapy (IMRT) is the most commonly used RT modality for esophageal cancer. However, there is increasing interest in the potential use of proton RT (PRT) for esophageal cancer owing to favorable dose distribution with reduction in integral dose delivered to the total body, heart, lungs, stomach, and bowel.3-6 Preliminary retrospective data suggest that PRT is a feasible and safe technique for esophageal cancer, with suggestion of reduced treatmentrelated adverse effects and favorable survival outcomes.7,8 However, there are no studies comparing patient-reported health-related quality of life (HRQoL) for patients with esophageal cancer treated with PRT versus XRT. The purpose of this analysis was to compare HRQoL during CRT for patients with esophageal cancer receiving PRT or XRT. HRQoL was assessed using the Functional Assessment of Cancer Therapy-Esophageal (FACT-E) questionnaire, a validated instrument consisting of a general questionnaire (FACT-G) and an esophageal cancer subscale (ECS).9,10 Owing to the more favorable dose distribution observed with PRT versus XRT for esophageal cancer, we hypothesized that PRT would be associated with less decline in HRQoL during CRT compared with XRT.
Methods and Materials Patient selection Starting in June 2015, patients treated with curativeintent RT at 2 campuses from 1 institution were approached by a research coordinator at the time of
simulation or first RT treatment for potential participation in a patient outcomes registry. For this institutional review boardeapproved prospective study, patients provided written consent for use of their data for research purposes. Patients were administered HRQoL studies on a portable electronic device at registration (simulation or first RT treatment) and during the last week of RT. For this analysis, inclusion criteria were esophageal or gastroesophageal junction carcinoma treated with curative-intent preoperative or definitive RT between June 2015 and February 2018. Exclusion criteria were prior esophagectomy or gastrectomy, prior thoracic RT, a palliative course of treatment (planned RT dose <40 Gy), receipt of both PRT and XRT, lack of concomitant chemotherapy, or no pretreatment questionnaire.
Treatment Patients received CRT in the preoperative or definitive setting. Pretreatment feeding tubes were uncommonly used (<10% of patients), and typically only for patients with complete esophageal obstruction. CRT was categorized as preoperative or definitive based on actual treatment received (ie, whether the patient actually underwent esophagectomy). Standard chemotherapy consisted of weekly intravenous carboplatin 50 mg/m2 and paclitaxel titrated to an area under the curve of 2. Decision on whether to use PRT or XRT was based primarily on insurance coverage and treatment scheduling availability. All patients underwent simulation with a 4-dimensional CT scan. For both modalities, the clinical tumor volume (CTV) included the primary tumor with a 3- to 4-cm mucosal expansion and a 1-cm radial expansion as well as involved regional lymph nodes with a 1-cm expansion. The CTV typically included the celiac lymph nodes for distal esophagus or gastroesophageal junction tumors and the supraclavicular lymph nodes for upper esophagus tumors. CTVs were typically modified to exclude uninvolved adjacent organs at risk, including the heart, lungs, bones, and bowel. For XRT, the planning target volume was a 5-mm uniform expansion of the CTV. For PRT, robust treatment planning was used with a goal to have at least 95% of the CTVs receiving 95% of the prescription dose with translational shifts of 5 mm in x, y, and z directions and range uncertainty of 3%. PRT was
Practical Radiation Oncology: --- 2019
Figure 1
Photon or proton RT for esophageal cancer
3
Example of photon (left) and proton (right) dose distributions for a sample patient.
administered using pencil beam spot scanning, typically with 2 posterior oblique fields. XRT typically consisted of volumetric modulated arc RT (70% of patients) or 3dimensional conformal RT (30% of patients). Examples of dosimetric features for PRT and XRT are depicted in Figure 1. For both PRT and XRT, daily image guidance consisted of orthogonal kilovoltage x-rays. The most common RT dose and fractionation regimens used were 50 Gy in 25 fractions or 50.4 Gy in 28 fractions. For patients receiving preoperative RT, 41.4 Gy in 23 fractions was also an option, which was more commonly used with XRT, especially if dose constraints for the heart (mean 26 Gy) or lung (mean 15Gy) could not be met with a dose of 50 Gy.
HRQoL questionnaires At baseline and at the end of CRT treatment, patients were asked to complete the FACT-E survey. The FACT-E contains the FACT-G and the ECS, which have been validated for evaluation of patient-reported outcomes in the context of esophageal cancer treatments.9,10 The FACT-G covers domains of physical well-being (PWB), social and family well-being (SWB), emotional wellbeing (EWB), and functional well-being (FWB) for adults with cancer.9,11 The ECS contains 17 items addressing symptoms and problems related to esophageal cancer. The FACT-E trial outcome index includes the PWB, FWB, and ECS subscales. The FACT-E takes approximately 10 minutes to complete. The FACT-E total score is the sum of the FACT-G scores and the ECS; higher scores indicate a better health-related QoL. The maximum FACT-E score is 176 points.
Statistics Baseline patient and treatment characteristics were compared between the XRT and PRT cohorts using the 2-sample t test or Fisher’s exact test. Change in HRQoL scores from baseline to end of treatment were assessed using the 2-sample t test. Univariable and multivariable linear regression analyses were used to assess for associations between patient and treatment characteristics and HRQoL score changes. For all statistical tests, a P value of <.05 was considered statistically significant. We performed sensitivity analyses, including patients who had completed the baseline questionnaire but not the end-of-CRT questionnaire. The patient’s chart was reviewed, and the end-of-treatment questionnaire was determined to be “missing at random” or “missing not at random.” Patients were determined to be “missing not at random” if documentation suggested clinical deterioration with acute grade 3 toxicity based on the Common Terminology Criteria for Adverse Events (v4.03), hospital admission, or a non-cancer-related event precluding patient participation. Patients were determined to be “missing at random” if the previous criteria were not met and documentation indicated that the patient would have been capable of filling out a questionnaire at the end of treatment. Two imputation models were designed to take missing posttreatment questionnaires into account. In both models, the patients with “missing at random” questionnaires were assigned a predicted score using a regression model with age, sex, and baseline score as covariates. In the first model, the patients with “missing not at random” questionnaires were attributed a score of the worst recorded score among patients with posttreatment
4
Practical Radiation Oncology: --- 2019
A. Garant et al Table 1
Patient and treatment characteristics for patients with a baseline and end-of-treatment FACT-E questionnaire
Age at diagnosis, y, mean (SD) Sex, no. (%) Female Male Tumor location, no. (%) Upper or middle Lower or gastroesophageal junction Histology, no. (%) Adenocarcinoma Squamous Other* Stage, no. (%) Iy II III IVz Not applicable* Therapy received, no. (%) Definitive Trimodality Concurrent chemotherapy, no. (%) Carboplatin/Paclitaxel Otherx Clinical target volume, cm3, mean (SD) Radiation therapy fractions, mean (SD) Radiation therapy dose received, no. (%) <5000 cGy 4140 cGy 5000 cGy 5000 cGy 5040 cGy
Total (n Z 125)
Proton (n Z 62)
Photon (n Z 63)
65.9 (10.6)
68.9 (9.9)
62.9 (10.5)
21 (16.8) 104 (83.2)
9 (14.5) 53 (85.5)
P Value .0014 .63
12 (19.0) 51 (81.0) 1.0
107 (85.6) 18 (15.4)
53 (85.5) 9 (14.5)
54 (85.7) 9 (14.3) .30
97 (77.6) 17 (13.6) 11 (8.8)
51 (82.3) 8 (12.9) 3 (4.8)
46 (73.0) 9 (14.3) 8 (12.7)
9 21 88 4
5 13 42 2
4 8 46 2
.66 (7.4) (17.2) (72.1) (3.3) 3
(8.1) (21.0) (67.7) (3.2) 0
(6.7) (13.3) (76.7) (3.3) 3 .14
49 (39.2) 76 (60.8)
20 (32.3) 42 (67.7)
29 (46.0) 34 (54.0)
120 (96.0) 5 (4.0) 611 (258) 25.2 (1.6)
61 (98.4) 1 (1.6) 553 (238) 24.9 (0.6)
59 (93.6) 4 (6.4) 668 (267) 25.4 (2.2)
25 (20.0)
4 (6.5) 2 58 (93.5) 58 1
21 (33.3) 19 42 (66.7) 20 20
.21
100 (80.0)
.013 .14 .0002
Abbreviations: FACT-E Z Functional Assessment of Cancer Therapy-Esophagus; SD Z standard deviation. * Localized neuroendocrine or small cell carcinoma. y Patients with stage I disease were medically inoperable or declined esophagectomy. z Patients with stage IV disease had limited metastatic spread to contiguous lymph nodes or organs and were treated with curative-intent radiation therapy and chemotherapy. x Three patients received cisplatin and etoposide, and 2 patients received fluorouracil and oxaliplatin.
data minus 1 point; in the second model, the patients with “missing not at random” questionnaires were attributed a score of 2 out of 4 points (50%) for each question.
Results Between June 2015 and February 2018, 189 patients met the study inclusion criteria and filled out a baseline questionnaire. In total, 125 patients completed both a baseline and end-of-treatment FACT-E. All patients received the planned RT dose except for 1 XRT patient who stopped after 34.2 Gy. The end-of-treatment survey was completed a median of 1 day (interquartile range [IQR], 0-3) before the last day of RT, with no significant difference observed between the PRT (median 1 day; IQR, 0-3) and XRT (median 1 day; IQR, 0-2) cohorts
(P Z .15 by KruskaleWallis test). Patient and treatment characteristics are summarized in Table 1. There were significant differences observed between PRT and XRT cohorts in regard to age, CTV, and prescription dose. Specifically, the PRT cohort (compared with the XRT cohort) had an older mean age, a smaller mean CTV, and a higher prescription dose. The mean heart dose was 7.0 Gy for the PRT cohort and 20.2 Gy for the XRT cohort (P < .0001). The mean lung dose was 3.6 Gy for the PRT cohort and 8.9 Gy for the XRT cohort (P < .0001). Results of the FACT-E surveys are summarized in Table 2. The mean FACT-E score was 136.3 points (standard deviation [SD] 21.0) at baseline. The baseline mean FACT-E score was similar for patients receiving preoperative versus definitive therapy (137.2 vs 135.0, P Z .57) and for patients receiving PRT versus XRT
Practical Radiation Oncology: --- 2019 Table 2
Photon or proton RT for esophageal cancer
5
FACT-E results for patients with baseline and end-of-treatment questionnaires (mean, standard deviation)
FACT-E total scores Baseline Post-treatment FACT-G total scores Baseline Post-treatment Differences between baseline and post-treatment scores FACT-E FACT-G FACT-E Trial Outcome Index Esophageal cancer subscale Emotional well-being subscale Functional well-being subscale Physical well-being subscale Social and family well-being subscale
Total (n Z 125)
Proton (n Z 62)
Photon (n Z 63)
P Value
136.3 (21.0) 119.6 (24.8)
136.7 (23.1) 124.0 (26.9)
135.9 (18.8) 115.3 (21.9)
.82 .048
87.0 (13.7) 77.4 (16.3)
87.2 (14.5) 80.3 (16.5)
86.9 (13.1) 74.5 (15.8)
.89 .047
e16.7 e9.7 e16.1 e7.0 e0.4 e3.1 6.0 e0.1
(19.8) (12.7) (17.2) (11.1) (4.4) (5.6) (5.4) (4.3)
e12.7 e6.9 e13.1 e5.8 0.0 e1.6 5.6 0.4
(19.9) (12.5) (17.4) (10.5) (4.4) (6.1) (5.1) (4.1)
e20.6 e12.4 e19.2 e8.3 e0.8 e4.6 e6.3 e0.6
(19.1) (12.3) (16.6) (11.7) (4.3) (4.7) (5.6) (4.3)
.026 .015 .048 .22 .29 .0032 .43 .20
Abbreviations: FACT-E Z Functional Assessment of Cancer Therapy-Esophagus; FACT-G Z Functional Assessment of Cancer Therapy-General.
(136.7 vs 135.9, P Z .82). The mean FACT-E score for the whole cohort was 119.6 points (SD 24.8) at the end of CRT, with mean change of e16.7 (SD 19.8). The greatest numerical declines were noted in the ECS (mean change of e7), PWB (mean change of e6), and FWB (mean change of e3.1), whereas there was no significant change in social and emotional well-being. In the univariable linear regression model (Table 3), less mean decline in FACT-E was observed for PRT versus XRT (-12.7 vs -20.6, P Z .026) and for trimodality therapy versus definitive therapy (-13.0 vs -22.5, P Z .008). There was also significantly less mean decline with PRT versus XRT for the FACT-G (-6.9 vs -12.4, P Z .015) and the FACT-E Trial Outcome Index (-13.1 vs -19.2, P Z .048). For all survey subscales, the median decline was numerically less in the PRT versus XRT cohort, although this was only statistically different for the FWB Subscale (-1.6 vs -4.6, P Z .0032). In the multivariable linear regression model, PRT (vs XRT, P Z .026) and trimodality therapy (vs definitive therapy, P Z .004) were associated with less decline in FACT-E. Sensitivity analyses were performed on the entire cohort of 189 patients with a baseline FACT-E survey, including the 64 patients (34%) who did not complete a posttreatment survey (patient and treatment characteristics are outlined in Table E1, available online at https://doi. org/10.1016/j.prro.2019.07.003). This included 20 patients who received PRT and 44 patients who received XRT. Among the 64 missing posttreatment surveys, 48 surveys were determined to be missing at random and 16 were missing not at random. The percentage of surveys missing at random versus not at random was not different for the PRT versus XRT cohorts. When imputing data for missing questionnaires under several scenarios, the
different declines in scores of PRT and XRT cohorts remained significant. In the scenario where all nonrandomly missing questionnaires were imputed with a posttreatment score of the worst recorded score in the database minus 1, the XRT versus PRT mean score declines were -23.3 vs -16.0, respectively (P Z .022). In the scenario where all nonrandomly missing questionnaires were imputed with a posttreatment score of 50%, the XRT versus PRT median score declines were -20.7 and -13.9, respectively (P Z .015). These changes in HRQoL scores are depicted in Table 4 and Figure 2.
Discussion In this analysis of patients with esophageal cancer receiving curative intent neoadjuvant or definitive CRT, we observed that HRQoL declined significantly from pretreatment to the end of CRT, consistent with previous observations.12 Importantly, RT modality was associated with decline in HRQoL, with significantly less decline in patients who received PRT versus XRT. This is one of the largest studies to date assessing patient-reported HRQoL during CRT for esophageal cancer. To our knowledge, this is the only study thus far assessing for association between RT modality and HRQoL during CRT for esophageal cancer. Additionally, this is one of few studies comparing HRQoL with XRT versus PRT for any type of cancer.13 Strengths of this study include the relatively large cohort of patients with esophageal cancer treated with CRT over a relatively short time interval; thus, treatment regimens were relatively homogenous, minimizing risk of bias. A vast majority of patients with esophageal cancer treated at our
6
Practical Radiation Oncology: --- 2019
A. Garant et al Table 3
Univariable and multivariable analyses of variables potentially associated with FACT-E change
Variable
Univariable
Age, y Male (vs female) Adenocarcinoma (vs squamous cell) stage III or IV (vs I or II) Trimodality (vs definitive) CTV (per 100 cm3) RT dose 5000 cGy (vs < 5000 cGy) Proton (vs photon)
Multivariable
Estimate, points
Standard error, points
P Value
Estimate, points
Standard error, points
P Value
0.12 3.22 1.94
0.17 4.76 5.13
.49 .50 .71
0.14 1.49 0.19
0.18 5.62 5.3
.43 .79 .97
5.69 9.55 0.090 4.27
4.17 3.55 0.69 4.44
.18 .008 .90 .34
4.40 11.07 1.22 0.47
4.41 3.74 0.83 4.78
.32 .004 .14 .92
7.89
3.49
.026
9.21
4.07
.026
Abbreviations: CTV Z clinical target volume; FACT-E Z Functional Assessment of Cancer Therapy-Esophagus; RT Z radiation therapy.
institutions (95%) had competed baseline HRQoL, minimizing risk of selection bias. To our knowledge, there have not been any studies that have defined minimally important difference or clinically meaningful change in the FACT-E in patients with esophageal cancer. In a study of patients with gastric cancer, a mean change in FACT-G score of 3.9, 7.8, and 11.8 points represented a small, medium, or large change in QoL, respectively.14 A mean change in the FACT-Ga (including the FACT-G and gastric cancer subscale) score of 7.5, 15.1, and 22.6 points represented a small, medium, or large change in QoL, respectively. Other studies have defined a one-third change in standard deviation in FACT scores as clinically meaningful change.15,16 Using data derived from the present study, a one-third standard deviation would correspond to approximately 5 points for the FACT-G and approximately 7.5 points for the FACT-E. Therefore, the observed statistically significant differences of 5.5 points for the FACT-G and 7.9 points for the FACT-E between XRT and PRT patients in our study likely represents a small but clinically meaningful change in QOL. There are limitations to our analysis. Given that this study was retrospective, there are potential selection biases and confounders that are difficult to account for in a nonrandomized setting. Specifically, there were significant differences between the PRT and XRT cohorts with
Table 4
respect to patient age, RT prescription dose, and CTV. The discrepancy in age between the PRT and XRT cohorts relates to the fact that during the study period, most private insurance carriers did not cover PRT for esophageal cancer, whereas our Medicare carrier did. However, patient age, RT prescription dose, and CTV were not associated with a change in FACT-E score during CRT, suggesting that imbalances in the covariates between the PRT and XRT cohorts were unlikely to explain the observed change in FACT-E. Similarly, a previous study did not demonstrate an association between patient age and change in QOL during CRT for esophageal cancer.17 Furthermore, the baseline FACT-E scores were similar for the PRT and XRT cohorts (136.7 vs 135.9, P Z .82), arguing against clinically significant imbalance in the pretreatment patient or disease characteristics. Sixty-four patients (34%) were excluded from the primary analysis because they did not complete an end-of-treatment HRQoL questionnaire. Unfortunately, missing data are a common occurrence in QoL research in patients with cancer.18-20 The rate of missing data at the end of treatment in our study is similar to that observed in recent prospective studies of QoL during CRT for esophageal cancer, in which end-of-treatment surveys were missing in 18% to 41% of patients.21-23 We performed sensitivity analyses to account for missing end-of-treatment surveys using common imputation methods, which also showed a
FACT-E results including patients with imputed scores for missing questionnaires, mean (standard deviation)
Differences between baseline and post-treatment scores
Total (n Z 189)
Proton (n Z 82)
Photon (n Z 107)
P Value
FACT-E, imputation method 1 FACT-E, imputation method 2 FACT-G, imputation method 1 FACT-G, imputation method 2
20.1 17.8 13.4 10.5
16.0 13.9 10.4 7.9
23.3 20.7 15.7 12.5
.022 .015 .036 .011
(21.8) (19.0) (17.2) (12.3)
(22.4) (20.7) (16.3) (12.7)
(20.9) (17.1) (17.6) (11.8)
Abbreviations: FACT-E Z Functional Assessment of Cancer Therapy-Esophagus; FACT-G Z Functional Assessment of Cancer Therapy-General.
Practical Radiation Oncology: --- 2019
Photon or proton RT for esophageal cancer
status.26-28 Djarv et al studied QoL changes in 132 patients with esophageal or gastric cancer who had received curative-intent CRT or surgery.29 QoL was assessed using the European Organization for Research and Treatment of Cancer QLQ-C30 at baseline and after 6 months. In a Cox proportional hazard analyses adjusted for age, sex, performance status, tumor stage, histology, disease site, baseline HRQoL, and treatment, better recovery of physical function at 6 months was associated with better survival; a 10-point increase in problems with pain and fatigue was associated with a higher risk of death. Further evaluation is needed to better establish the associations between HRQoL and oncologic outcomes.
175
150
FACT E Score
125
100
75
50
25
0
7
Radiation type
Conclusions
Photon Proton Baseline
Post−Tx
Patients with both surveys
Baseline
Post−Tx
All Patients, Imputation method 1
Baseline
Post−Tx
All Patients, Imputation method 2
Figure 2 FACT-E score changes for proton versus photon cohorts (dot Z mean; bracket Z standard deviation).
significant difference in FACT-E scores for the PRT versus XRT cohorts. We hypothesize that the observation of less decline in HRQoL with PRT versus XRT is related to the more favorable dose distribution of PRT versus XRT for esophageal cancer, as illustrated in Figure 1. Specifically, PRT (compared with XRT) reduces integral dose to most organs at risk including the heart, lungs, stomach, bowel, liver, and kidneys.3-6 Emerging clinical data suggest PRT may be associated with a lower incidence of treatmentrelated adverse effects including severe lymphopenia and favorable survival outcomes compared with XRT.8,24,25 Reduction in dose to the stomach, bowel, circulating blood cells, and total body would be anticipated to reduce acute treatment-related nausea, anorexia, and fatigue, all components that affect HRQoL. An ongoing phase 2 randomized trial is comparing PRT versus XRT for esophageal cancer (NCT01512589). Endpoints include toxicity burden, HRQoL, and survival. Additionally, a phase 3 randomized controlled trial of PRT versus XRT for esophageal cancer is now open through NRG Oncology (NRG-GI006, NCT03801876). These prospective trials will provide higher-quality data comparing adverse events, HRQoL, and survival with PRT versus XRT. Several investigators have found that HRQoL scores have prognostic implications for patients undergoing therapy for esophageal cancer. In a pooled analysis of 128 patients in 4 prospective, nonrandomized studies of patients with esophageal cancer who received trimodality or definitive therapy, Kidane et al found that pretreatment FACT-E scores had better discrimination for survival than Eastern Cooperative Oncology Group performance
In this observational study of patients with esophageal cancer undergoing curative-intent CRT, use of PRT was associated with significantly less decline in FACT-E scores during treatment compared with XRT. Although the integration HRQoL research is challenging in the outpatient clinic workflow, randomized trials are needed to confirm our findings, and further studies are needed to determine whether changes in HRQoL during CRT are associated with long-term QoL, toxicity, and survival.
Supplementary Data Supplementary material for this article can be found at https://doi.org/10.1016/j.prro.2019.07.003.
References 1. Shapiro J, van Lanschot JJB, Hulshof MCCM, 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. 2. Cooper JS, Guo MD, 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. 3. Chuong MD, Hallemeier CL, Jabbour SK, et al. Improving outcomes for esophageal cancer using proton beam therapy. Int J Radiat Oncol Biol Phys. 2016;95:488-497. 4. Lin SH, Hallemeier CL, Chuong M. Proton beam therapy for the treatment of esophageal cancer. Chin Clin Oncol. 2016;5:53. 5. Lin SH, Komaki R, Liao Z, et al. Proton beam therapy and concurrent chemotherapy for esophageal cancer. Int J Radiat Oncol Biol Phys. 2012;83:e345-351. 6. Xi M, Lin SH. Recent advances in intensity modulated radiotherapy and proton therapy for esophageal cancer. Expert Rev Anticancer Ther. 2017;17:635-646. 7. Lin SH, Merrell KW, Shen J, et al. Multi-institutional analysis of radiation modality use and postoperative outcomes of neoadjuvant chemoradiation for esophageal cancer. Radiother Oncol. 2017;123: 376-381. 8. Xi M, Xu C, Liao Z, et al. Comparative outcomes after definitive chemoradiotherapy using proton beam therapy versus intensity
8
9.
10.
11.
12.
13.
14.
15.
16.
17.
18.
19.
A. Garant et al modulated radiation therapy for esophageal cancer: A retrospective, single-institutional analysis. Int J Radiat Oncol Biol Phys. 2017;99: 667-676. Cella DF, Tulsky DS, Gray G, et al. The Functional Assessment of Cancer Therapy scale: development and validation of the general measure. J Clin Oncol. 1993;11:570-579. Darling G, Eton DT, Sulman J, Casson AG, Celia D. Validation of the functional assessment of cancer therapy esophageal cancer subscale. Cancer. 2006;107:854-863. Overcash J, Extermann M, Parr J, Perry J, Balducci L. Validity and reliability of the FACT-G scale for use in the older person with cancer. Am J Clin Oncol. 2001;24:591-596. Revicki D, Hays RD, Cella D, Sloan J. Recommended methods for determining responsiveness and minimally important differences for patient-reported outcomes. J Clin Epidemiol. 2008;61:102-109. Verma V, Simone CB, Mishra MV. Quality of life and patientreported outcomes following proton radiation therapy: A systematic review. J Natl Cancer Inst. 2017;110:341-353. Garland SN, Pelletier G, Lawe A, et al. Prospective evaluation of the reliability, validity, and minimally important difference of the functional assessment of cancer therapy-gastric (FACT-Ga) qualityof-life instrument. Cancer. 2011;117:1302-1312. Steel JL, Eton DT, Cella D, Olek MC, Carr BI. Clinically meaningful changes in health-related quality of life in patients diagnosed with hepatobiliary carcinoma. Ann Oncol. 2006;17:304-312. Cella D, Eton DT, Fairclough DL, et al. What is a clinically meaningful change on the Functional Assessment of Cancer Therapy-Lung (FACT-L) Questionnaire? Results from Eastern Cooperative Oncology Group (ECOG) Study 5592. J Clin Epidemiol. 2002;55:285-295. Trudel JG, Sulman J, Atenafu EG, Kidane B, Darling GE. Longitudinal evaluation of trial outcome index scores in patients with esophageal cancer. Ann Thorac Surg. 2016;102:269-275. Donaldson GW, Moinpour CM. Learning to live with missing quality-of-life data in advanced-stage disease trials. J Clin Oncol. 2005;23:7380-7384. Macefield RC, Jacobs M, Blencowe NS, et al. The case for a HRQL core outcome set: Outcome reporting bias in oesophageal cancer studies. Trials. 2011;12:A77.
Practical Radiation Oncology: --- 2019 20. Ter Veer E, van Kleef JJ, Sprangers MAG, Haj Mohammad N, van Oijen MGH, van Laarhoven HWM. Reporting of health-related quality of life in randomized controlled trials involving palliative systemic therapy for esophagogastric cancer: A systematic review. Gastric Cancer. 2018;21:1-13. 21. Noordman BJ, Verdam MGE, Lagarde SM, et al. Effect of neoadjuvant chemoradiotherapy on health-related quality of life in esophageal or junctional cancer: Results from the randomized CROSS trial. J Clin Oncol. 2018;36:268-275. 22. Rees J, Hurt CN, Gollins S, et al. Patient-reported outcomes during and after definitive chemoradiotherapy for oesophageal cancer. Br J Cancer. 2015;113:603. 23. Bascoul-Mollevi C, Gourgou S, Galais MP, et al. Health-related quality of life results from the PRODIGE 5/ACCORD 17 randomised trial of FOLFOX versus fluorouracilecisplatin regimen in oesophageal cancer. Eur J Cancer. 2017;84:239-249. 24. Davuluri R, Jiang W, Fang P, et al. Lymphocyte nadir and esophageal cancer survival outcomes after chemoradiation therapy. Int J Radiat Oncol Biol Phys. 2017;99:128-135. 25. Wang J, Wei C, Tucker SL, et al. Predictors of postoperative complications after trimodality therapy for esophageal cancer. Int J Radiat Oncol Biol Phys. 2013;86:885-891. 26. Kidane B, Sulman J, Xu W, et al. Pretreatment quality-of-life score is a better discriminator of oesophageal cancer survival than performance status. Eur J Cardiothorac Surg. 2017;51: 148-154. 27. Kidane B, Sulman J, Xu W, et al. Patient-derived score better discriminator of overall survival than clinician-assigned ECOG status. Interact Cardiovasc Thorac Surg. 2015;21:S33S34. 28. Kidane B, Sulman J, Xu W, et al. Baseline measure of healthrelated quality of life (Functional Assessment of Cancer TherapyEsophagus) is associated with overall survival in patients with esophageal cancer. J Thorac Cardiovasc Surg. 2016;151:15711580. 29. Djarv T, Metcalfe C, Avery KN, Lagergren P, Blazeby JM. Prognostic value of changes in health-related quality of life scores during curative treatment for esophagogastric cancer. J Clin Oncol. 2010; 28:1666-1670.