Predicting Esophagitis After Chemoradiation Therapy
for Non-Small Cell Lung Cancer: An Individual Patient
Data Meta-Analysis
Citation for published version (APA):
Palma, D. A., Senan, S., Oberije, C., Belderbos, J., de Dios, N. R., Bradley, J. D., Barriger, R. B., Moreno-
Jimenez, M., Kim, T. H., Ramella, S., Everitt, S., Rengan, R., Marks, L. B., De Ruyck, K., Warner, A., &
Rodrigues, G. (2013). Predicting Esophagitis After Chemoradiation Therapy for Non-Small Cell Lung
Cancer: An Individual Patient Data Meta-Analysis. International Journal of Radiation Oncology Biology
Physics, 87(4), 690-696. https://doi.org/10.1016/j.ijrobp.2013.07.029
Document status and date:
Published: 15/11/2013
DOI:
10.1016/j.ijrobp.2013.07.029
Document Version:
Publisher's PDF, also known as Version of record
Document license:
Taverne
Please check the document version of this publication:
• A submitted manuscript is the version of the article upon submission and before peer-review. There can
be important differences between the submitted version and the official published version of record.
People interested in the research are advised to contact the author for the final version of the publication,
or visit the DOI to the publisher's website.
• The final author version and the galley proof are versions of the publication after peer review.
• The final published version features the final layout of the paper including the volume, issue and page
numbers.
Link to publication
General rights
Copyright and moral rights for the publications made accessible in the public portal are retained by the authors and/or other copyright
owners and it is a condition of accessing publications that users recognise and abide by the legal requirements associated with these
rights.
• Users may download and print one copy of any publication from the public portal for the purpose of private study or research.
• You may not further distribute the material or use it for any profit-making activity or commercial gain
• You may freely distribute the URL identifying the publication in the public portal.
If the publication is distributed under the terms of Article 25fa of the Dutch Copyright Act, indicated by the “Taverne” license above,
please follow below link for the End User Agreement:
www.umlib.nl/taverne-license
Take down policy
If you believe that this document breaches copyright please contact us at:
repository@maastrichtuniversity.nl
providing details and we will investigate your claim.
Download date: 09 Aug. 2022
Clinical Investigation: Thoracic Cancer
Predicting Esophagitis After Chemoradiation Therapy for
Non-Small Cell Lung Cancer: An Individual Patient Data
Meta-Analysis
David A. Palma, MD, MSc, PhD,
*
Suresh Senan, MRCP, FRCR, PhD,
y
Cary Oberije, MSc, PhD,
z
Jose Belderbos, MD, PhD,
x
Nu
´
ria Rodrı
´
guez de Dios, MD PhD,
k
Jeffrey D. Bradley, MD,
{
R. Bryan Barriger, MD,**
Marta Moreno-Jime
´
nez, MD, PhD,
yy
Tae Hyun Kim, MD,
zz
Sara Ramella, MD,
xx
Sarah Everitt, PhD,
kk
Ramesh Rengan, MD, PhD,
{{
Lawrence B. Marks, MD,***
Kim De Ruyck, PhD,
yyy
Andrew Warner, MSc,
*
and George Rodrigues, MD, MSc
*
,zzz
*Department of Radiation Oncology, London Regional Cancer Program, London, Ontario, Canada,
y
Department of Radiation
Oncology, VU University Medical Center, Amsterdam, The Netherlands,
z
Department of Radiation Oncology (MAASTRO
Clinic), GROW e School for Oncology and Developmental Biology, Maastricht University Medical Centre, Maastricht, The
Netherlands,
x
Department of Radiation Oncology, The Netherlands Cancer Institute e Antoni van Leeuwenhoek Hospital,
Amsterdam, The Netherlands,
k
Department of Radiation Oncology, Parc de Salut Mar, Barcelona, Universidad Pompeu
Fabra, Barcelona, Spain,
{
Department of Radiation Oncology, Washington University School of Medicine, St. Louis,
Missouri, USA, **Department of Radiation Oncology, Indiana University School of Medicine, Indianapolis, Indiana, USA,
yy
Department of Oncology, Radiation Oncology Division, Clı
´
nica Universidad de Navarra, University of Navarra, Pamplona,
Spain,
zz
Center for Proton Therapy, Research Institute and Hospital, National Cancer Center, Goyang, Gyeonggi, Korea,
xx
Division of Radiation Oncology, Campus Bio-Medico University, Rome, Italy,
kk
Radiation Therapy Services, Peter
MacCallum Cancer Centre, Melbourne, Australia and Department of Medical Imaging and Radiation Sciences, Faculty of
Medicine, Nursing & Health Sciences, Monash University, Australia,
{{
Department of Radiation Oncology, University of
Pennsylvania, Philadelphia, Pennsylvania, USA, ***Department of Radiation Oncology, University of North Carolina,
Chapel Hill, North Carolina, USA,
yyy
Department of Basic Medical Sciences, Ghent University, Ghent, Belgium, and
zzz
Department of Epidemiology and Biostatistics, University of Western Ontario, London, Ontario, Canada
Received Mar 18, 2013, and in revised form Jun 25, 2013. Accepted for publication Jul 25, 2013
Reprint requests to: David A. Palma, MD, MSc, PhD, Department of
Radiation Oncology, London Regional Cancer Program, 790 Commis-
sioners Rd E, London, Ontario, Canada, N6A4L6. Tel: (þ1) 519-685-
8500; E-mail: david.palma@uwo.ca
Accepted for oral presentations at the Canadian Association of Radia-
tion Oncologists’ Annual Meeting 2013 and the 55th Annual Meeting of
the American Society for Radiation Oncology (ASTRO), Atlanta, GA,
September 22-25, 2013.
Supported by a clinician-scientist grant from the Ontario Institute for
Cancer Research (to D.A.P.), and by National Institutes of Health Grant
CA69579 (to L.B.M.).
Conflict of interest: Dr Senan has received research funding from
Sanofi-Aventis and is a member of the trial management group (not
reimbursed) for the phase 3 PROCLAIM study evaluating concurrent
chemoradiation therapy schemes for stage III lung cancer, which is
sponsored by Eli Lilly. Dr Barriger holds a leadership position with D3
Oncology Solutions, is a member of the Via Oncology Pathways Physician
Advisory Committee, and is the Radiation Oncology Esophageal and Lung
(Small Cell and Non-Small Cell) Committee Co-Chair. The authors report
no other conflict of interest.
AcknowledgmentsdThe authors thank Gabriel Bolt for assistance with
literature review, and Drs Achilles Fakiris, Joke Bakker, Daniel Mullen,
Ellen van Reij, Michael Lawrence, Eric Xanthopoulos, Shiva Das, Mary
Duffy, Philippe Lambin, and David Ball for assistance with provision of
data.
Int J Radiation Oncol Biol Phys, Vol. 87, No. 4, pp. 690e696, 2013
0360-3016/$ - see front matter Ó 2013 Elsevier Inc. All rights reserved.
http://dx.doi.org/10.1016/j.ijrobp.2013.07.029
Radiation Oncology
International Journal of
biology physics
www.redjournal.org
Summary
This meta-analysis examined
predictors of radiation
esophagitis (RE) for patients
receiving concurrent chemo-
radiation therapy for non-
small cell lung cancer.
Clinically significant RE is
common, but life-threatening
complications occur in <1%
of patients. The esophageal
volume receiving 60 Gy
(V60) is the best predictor of
RE, and additional variables
did not improve predictive
ability. A V60 <0.07% is
associated with a low risk of
RE, whereas a V60 17%
confers the highest risk.
Purpose: Concurrent chemoradiation therapy (CCRT) improves survival compared with sequen-
tial treatment for locally advanced non-small cell lung cancer, but it increases toxicity, particu-
larly radiation esophagitis (RE). Validated predictors of RE for clinical use are lacking. We
performed an individual-patient-data meta-analysis to determine factors predictive of clinically
significant RE.
Methods and Materials: After a systematic review of the literature, data were obtained on 1082
patients who underwent CCRT, including patients from Europe, North America, Asia, and
Australia. Patients were randomly divided into training and validation sets (2/3 vs 1/3 of
patients). Factors predictive of RE (grade 2 and grade 3) were assessed using logistic
modeling, with the concordance statistic (c statistic) used to evaluate the performance of each
model.
Results: The median radiation therapy dose delivered was 65 Gy, and the median follow-up
time was 2.1 years. Most patients (91%) received platinum-containing CCRT regimens. The
development of RE was common, scored as grade 2 in 348 patients (32.2%), grade 3 in 185
(17.1%), and grade 4 in 10 (0.9%). There were no RE-related deaths. On univariable analysis
using the training set, several baseline factors were statistically predictive of RE (P<.05), but
only dosimetric factors had good discrimination scores (c > .60). On multivariable analysis,
the esophageal volume receiving 60 Gy (V60) alone emerged as the best predictor of grade
2 and grade 3 RE, with good calibration and discrimination. Recursive partitioning identified
3 risk groups: low (V60 <0.07%), intermediate (V60 0.07% to 16.99%), and high (V60 17%).
With use of the validation set, the predictive model performed inferiorly for the grade 2
endpoint (c Z .58) but performed well for the grade 3 endpoint (c Z .66).
Conclusions: Clinically significant RE is common, but life-threatening complications occur in
<1% of patients. Although several factors are statistically predictive of RE, the V60 alone
provides the best predictive ability. Efforts to reduce the V60 should be prioritized, with further
research needed to identify and validate new predictive factors. Ó 2013 Elsevier Inc.
Introduction
For patients with locally advanced non-small cell lung cancer
(NSCLC), the addition of concurrent chemotherapy to radiation
therapy provides the best chance of cure, achieving an absolute
improvement in survival of 10% at 2 years in comparison with
sequential chemotherapy and radiation (1). However, this
improvement in survival comes at a cost: patients receiving
concurrent chemoradiation therapy (CCRT) have an approxi-
mately 5-fold increase in the risk of acute radiation esophagitis
(RE) compared with patients receiving sequential treatment, and
in 1 randomized trial, 21% of patients on the CCRT arm had
treatment discontinued because of severe RE (1). Symptoms of
acute RE include odynophagia, dysphagia, and retrosternal pain,
in some cases resulting in weight loss and requiring analgesics,
intravenous fluids, hyperalimentation, the insertion of a percuta-
neous gastrostostomy tube, or some combination of these (2, 3).
The ability to accurately predict RE might facilitate strategies
to mitigate risk and thus improve the therapeutic ratio. To that end,
numerous studies have attempted to assess predictors of RE,
examining patient and tumor characteristics, dosimetric factors,
and the use of chemotherapy (2). However, such studies have not
yet resulted in the widespread adoption of any prediction model,
partly because of the heterogeneity of reported results across
studies, the lack of external validation, and the inclusion of
patients treated with older approaches (eg, sequential chemo-
therapy and radiation therapy) that may not be applicable to
modern practice. One of the largest previous studies developed
a practical prediction model (www.predictcancer.org) that was
validated on 3 separate cohorts but included patients with small
cell lung cancer and NSCLC, many of whom did not receive
concurrent chemotherapy (4).
Recent guidelines for assessing normal tissue toxicity risk
recommended the pooling of individual patient data to undertake
meta-analyses in an effort to overcome some of the limitations of
previous research (5). As a result, the collaborative project named
S
ystematic analysis of toxicity after radical irradiation: pneumo-
nitis and e
sophagitis (STRIPE) was launched to determine
predictors of radiation pneumonitis and esophagitis in patients
receiving CCRT. The first substudy examining predictors of
radiation pneumonitis was recently reported (6). The goal of this
component of the STRIPE project was to create and validate
a predictive model for RE in a population of patients with locally
advanced NSCLC receiving curative-intent, modern CCRT.
Methods and Materials
A systematic review was conducted using MEDLINE to identify
articles published between 1993 and January 2011 reporting on
dosimetric predictors of pneumonitis and RE. The full search
strategy, including electronic searches and hand searches for data
published in abstract form or unpublished, has been reported
previously (6). Authors were contacted and invited to submit
datasets (prospective or retrospective) with individual patient data.
Institutional Research Ethics Board approval was obtained.
Individual patients were included in the analysis if they were
treated for NSCLC with curative-intent CCRT, either
3-dimensional conformal radiation therapy (3D-CRT) or intensity
modulated radiation therapy (IMRT). Exclusion criteria included
stage IV disease, palliative-intent treatment, use of 2-dimensional
radiation therapy planning, small cell histology, surgical resection,
lack of esophagitis grade, and patients with no esophagitis dose
metrics available. Induction or adjuvant CCRT was permitted.
Patients who discontinued treatment because of toxicity (and
Volume 87 Number 4 2013 STRIPE esophagitis meta-analysis 691
therefore did not receive the intended dose) were included. All
centers reported contouring the esophagus from cricoid to
gastroesophageal junction as standard practice, and 1 center
routinely used contrast medium to assist in esophageal
delineation.
The 2 endpoints of this study were: (1) severe acute RE,
defined as grade 3 or higher according to the Common Termi-
nology Criteria for Adverse Events version 3 or 4, or grade 3 or
higher on the Radiation Therapy and Oncology Group (RTOG)
scale; and (2) moderate acute RE, defined as grade 2 or higher on
the same scales, which has been previously defined as a clinically
significant measure across multiple scoring scales (3). Patients
recorded as having RE during or after RT were recorded as having
an event.
Statistical analysis
A random number generator was used to separate patients into
a training set (nZ722, 2/3 of 1082) and a validation set (nZ360,
1/3 of 1082) without stratification. Descriptive statistics were
generated for baseline patient, tumor, and treatment characteris-
tics. Using the training set, univariable logistic regression analysis
was performed for each of the eligible available factors to identify
significant predictors of esophagitis grade 2 and grade 3. Due
to the fact that patients with missing data fields are excluded from
any regression analysis using those fields, it was required that any
predictor was available on at least 200 patients to be assessed as
a predictor. Imputation was not used because the data were
missing in a nonrandom fashion. By use of univariable analysis, P
values and concordance statistics (c statistics) were used to
quantify degree of association between each of the factors and
esophagitis endpoints. The c statistic indicates the discrimination
of the model (ie, the ability to distinguish low-risk from high-risk
study participants), and it ranges from .5 to 1, with .5 indicating
a useless test (no better than chance) and 1 indicating a perfect test
(7). For any 2 randomly selected patients of whom 1 has the
outcome of interest (RE) and 1 does not, the c statistic denotes
how often the model identifies the high-risk patient correctly.
Multivariable logistic regression models were then generated by
the use of forward-stepwise selection procedures, with factors
significant on univariable analysis eligible for inclusion in the
multivariable model. Factors remained included in the model if they
met 3 criteria: the factor was a statistically significant predictor, the
factor increased the discrimination of the model (denoted by
a higher c statistic), and the model was well calibrated, indicating
that the predicted risk for an individual patient is similar to the actual
risk, as judged by the Hosmer-Lemeshow test (7).
Recursive partitioning analysis (RPA) was performed incor-
porating significant predictors of esophagitis identified from
multivariable logistic regression as described previously (6) and
was used to create risk groups. RPA cutoffs were rounded to
increase clinical utility.
Once the analysis was complete on the training set, the
performance of the model and the RPA were evaluated by use of
the validation set. Survival estimates were calculated with the
Kaplan-Meier method, and the median follow-up time was
calculated with the reverse Kaplan-Meier method. Statistical
analysis was performed with SAS, version 9.2 (Cary, NC), and R,
version 2.15.2 (Vienna, Austria, for recursive partitioning analysis
only), with the use of 2-sided statistical testing at the .05 signif-
icance level. For the Hosmer-Lemeshow test of calibration, the
null hypothesis is that a model is well calibrated, such that P<.05
indicates poor calibration.
Results
Data were available on 1082 patients from 15 different sources,
12 previously reported in whole or in partd11 articles (4, 8-17)
and 1 abstract (18)d representing patients from North America,
Europe, Asia, and Australia (Table 1). Baseline clinical charac-
teristics are shown in Table 2. The median follow-up time was 26
months, and the median overall survival was 19 months. The
median total dose delivered was 65 Gy (range, 14-81.6 Gy), and
most patients received 2 Gy per fraction (85.2%). Numerous
chemotherapy regimens were used, with 91% of patients receiving
a platinum-containing regimen.
The development of RE was common, scored as grade 2 in 348
patients (32.2%), grade 3 in 185 patients (17.1%), and grade 4 in
10 patients (0.9%). There were no esophagitis-related deaths.
Table 1 Sources of individual patient data
Institution No. of patients
MAASTRO Clinic, Maastricht, The Netherlands 228
The Netherlands Cancer Institute e Antoni van Leeuwenhoek Hospital,
Amsterdam, Netherlands
136
Universidad Pompeu Fabra, Barcelona, Spain 100
Washington University School of Medicine, St. Louis, USA 80
Indiana University School of Medicine, Indianapolis, USA 73
Pulmonart Multicenter Study 71
University of Navarra, Pamplona, Spain 67
National Cancer Center, Goyang, Korea 65
London Regional Cancer Program 64
Campus Bio-Medico University, Rome, Italy 63
VU University Medical Center, Amsterdam, Netherlands 58
Peter MacCallum Cancer Centre, Melbourne, Australia 42
University of Pennsylvania, Philadelphia, USA 18
Duke University, Durham, USA 11
Ghent University Hospital, Ghent, Belgium 6
Total 1082
Palma et al. International Journal of Radiation Oncology Biology Physics692
Table 2 Baseline clinical characteristics
Characteristic N with data available at individual patient level Median (range) or n (%)
Age 1081 63.00 (24.53, 85.00)
Sex 1082
Male 750 (69.3)
Female 332 (30.7)
Histology 865
Adenocarcinoma 234 (27.1)
Squamous 288 (33.3)
Large cell 153 (17.7)
Other/NOS /undifferentiated 190 (22.0)
Stage
*
954
I 8 (0.8)
II 38 (4.0)
III 908 (95.2)
T stage 955
T0 5 (0.5)
T1 97 (10.2)
T2 302 (31.6)
T3 208 (21.8)
T4 316 (33.1)
TX 27 (2.8)
N stage 955
N0 121 (12.7)
N1 57 (6.0)
N2 522 (54.7)
N3 244 (25.6)
NX 11 (1.2)
Smoking history 538
Yes, current or former 504 (93.7)
Never 34 (6.3)
Performance status 709
Good (ECOG 0-1 or KPS 70) 655 (92.4)
Poor (ECOG 2 or KPS 70) 54 (7.6)
Neoadjuvant chemotherapy 222 139 (62.6)
Type of concurrent chemotherapy 771
Platinum-based þ etoposide 208 (27.0)
Platinum-based þ taxane 210 (27.2)
Other platinum-based 287 (37.2)
Other/not specified 66 (8.6)
Max dose to esophagus (Gy) 933 65.40 (0.40, 81.90)
Mean dose to esophagus (Gy) 561 32.60 (1.90, 68.70)
Esophagus V5 (%) 269 64.28 (20.64, 100.00)
Esophagus V10 (%) 369 58.39 (5.00, 100.00)
Esophagus V15 (%) 206 59.42 (11.08, 99.60)
Esophagus V20 (%) 411 48.63 (0.00, 100.00)
Esophagus V25 (%) 248 49.54 (0.00, 97.99)
Esophagus V30 (%) 429 44.06 (0.00, 100.00)
Esophagus V35 (%) 348 45.19 (0.00, 100.00)
Esophagus V40 (%) 411 39.00 (0.00, 100.00)
Esophagus V45 (%) 478 38.14 (0.00, 100.00)
Esophagus V50 (%) 502 29.21 (0.00, 98.70)
Esophagus V55 (%) 489 22.26 (0.00, 95.00)
Esophagus V60 (%) 486 11.04 (0.00, 93.64)
Abbreviations: ECOG Z Eastern Cooperative Oncology Group; NOS Z not otherwise specified.
* All staging data refer to AJCC, sixth edition.
Volume 87 Number 4 2013 STRIPE esophagitis meta-analysis 693
