European Journal of Surgical Oncology 47 (2021) 2016e2022

Contents lists available at ScienceDirect

European Journal of Surgical Oncology

journal homepage: www.ejso.com

Patterns of recurrence following definitive chemoradiation for

patients with proximal esophageal cancer
J. de Vos-Geelen a, *, S.M.E. Geurts a, G.A.P. Nieuwenhuijzen b, F.E.M. Voncken c,
J.A. Bogers d, P.M. Braam e, C.T. Muijs f, M.A. de Jong g, N. Kasperts h, T. Rozema i,
G.J. Blom j, S.A.W. Bouwense k, L.B.J. Valkenburg-van Iersel a, P.M. Jeene l, m,
F.J.P. Hoebers n, 1, V.C.G. Tjan-Heijnen a, 1
a
Dept. of Internal Medicine, Division of Medical Oncology, GROW - School for Oncology and Developmental Biology, Maastricht University Medical Center,
P.O. box 5800, 6202, AZ, Maastricht, the Netherlands
b
Dept. of Surgery, Catharina Hospital Eindhoven, P.O. box 1350, 5602, ZA, Eindhoven, the Netherlands
c
Dept. of Radiation Oncology, Netherlands Cancer Institute - Antoni van Leeuwenhoek Hospital, P.O. box 90203, 1006, BE, Amsterdam, the Netherlands
d
Radiotherapiegroep Arnhem, P.O. box 60160, 6800, JD, Arnhem, the Netherlands
e
Dept. of Radiotherapy, RadboudUMC, P.O. box 9101, 6500, HB, Nijmegen, the Netherlands
f
Dept. of Radiation Oncology, University Medical Center Groningen, University of Groningen, P.O. box 11120, 9700, RB, Groningen, the Netherlands
g
Dept. of Clinical Oncology, Leiden University Medical Centre, P.O. box 9699, 2300, RC, Leiden, the Netherlands
h
Dept. of Radiotherapy, University Medical Center Utrecht, P.O. box 85500, 3508, GA, Utrecht, the Netherlands
i
Insituut Verbeeten, P.O. box 90120, 5000, LA, Tilburg, the Netherlands
j
Dept. of Radiation Oncology, Amsterdam University Medical Centers, VU University, P.O. box 7057, 1007, MB, Amsterdam, the Netherlands
k
Dept. of Surgery, Maastricht University Medical Center, P.O. box 5800, 6202, AZ, Maastricht, the Netherlands
l
Dept. of Radiotherapy, Amsterdam University Medical Centers, University of Amsterdam, P.O. box 22660, 1100, DD, Amsterdam, the Netherlands
m
Radiotherapiegroep Deventer, P.O. box 123, 7400, AC, Deventer, the Netherlands
n
Dept. of Radiation Oncology (MAASTRO), GROW e School for Oncology and Developmental Biology, Maastricht University Medical Centre, P.O. box 3035,
6202, NA, Maastricht, the Netherlands

a r t i c l e i n f o a b s t r a c t

Article history: Introduction: The aim of this retrospective study was to determine the patterns of recurrence and overall
Received 14 October 2020 survival (OS) in patients achieving clinical complete response after treatment with definitive chemo-
Received in revised form radiation (CRT) for proximal esophageal cancer.
20 December 2020
Materials and methods: Patients with proximal esophageal cancer treated with CRT between 2004 and
Accepted 1 February 2021
Available online 5 February 2021
2014 in 11 centers in the Netherlands were included. OS and progression-free survival (PFS) were
calculated using the Kaplan-Meier method. Cumulative incidence of first recurrence (locoregional or
distant) and locoregional recurrence (LRR) were assessed using competing risk analyses.
Keywords:
Esophagus
Results: In 197 of the 200 identified patients, response was evaluated, 133 (68%) showed a complete
Cervical response. In complete responders, median OS, three-year OS, and PFS were 45.0 months (95% CI 34.8
Upper thoracic e61.5 months), 58% (95% CI 48e66), and 49% (95% CI 40e57), respectively. Three- and five-year risk of
Squamous cell cancer recurrence were respectively 40% (95% CI 31e48), and 45% (95% CI 36e54). Three- and five-year risk of
Chemoradiotherapy LRR were 26% (95% CI 19e33), and 30% (95% CI 22e38). Eight of 32 patients with an isolated LRR un-
Relapse derwent salvage surgery, with a median OS of 32.0 months (95% CI 6.8-not reached).
Conclusion: In patients with a complete response after definitive CRT for proximal esophageal cancer,
most recurrences were locoregional and developed within the first three years after CRT. These findings
suggest to shorten locoregional follow-up from five to three years.
© 2021 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY license
(http://creativecommons.org/licenses/by/4.0/).

* Corresponding author. Dept. of Internal Medicine, Div. of Medical Oncology, Maastricht University Medical Center, P.O. box 5800, 6202, AZ, Maastricht, the Netherlands.
E-mail address: judith.de.vos@mumc.nl (J. de Vos-Geelen).
1
These authors share senior authorship.

https://doi.org/10.1016/j.ejso.2021.02.001
0748-7983/© 2021 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/).
J. de Vos-Geelen, S.M.E. Geurts, G.A.P. Nieuwenhuijzen et al. European Journal of Surgical Oncology 47 (2021) 2016e2022

Introduction Data collection

Definitive chemoradiation (CRT) is the standard of care for pa- Patient demographics, tumor characteristics, treatment details,
tients with a proximal esophageal squamous cell cancer (SCC), and vital status were collected retrospectively from the medical
recommended by the National Comprehensive Cancer Network records, obtained by trained registry clerks. Data collection took
(NCCN), American Society of Clinical Oncology (ASCO), and Euro- place between April 2017 and May 2018.
pean Society for Medical Oncology (ESMO) guidelines [1e3].
Prognosis of proximal esophageal cancer (EC) is poor, providing
three-year overall survival (OS) rates in patients treated with CRT of Outcomes
about 35e45% [4e11]. Our previous retrospective cohort study
showed comparable OS between four different CRT regimens In the total group, we aimed to determine OS stratified by
including cisplatin or carboplatin and paclitaxel backbones and low response. In the complete responders, additional endpoints were
(
50.4 Gy) or high dose (>50.4 Gy) radiotherapy (RT) [12]. Long- progression-free survival (PFS), cumulative incidence of first
term survival is only achieved in complete responders to CRT recurrence (locoregional or distant), and cumulative incidence of
[7,9,13,14]. In the absence of clinical complete response (CR), none locoregional recurrence (LRR) as first event. Furthermore, we
of the patients were alive at three years following CRT [7,9], examined whether potential prognostic factors (age, sex, WHO
whereas three-year OS rates were shown to be approximately 25% performance status, comorbidity, clinical lymph node (cN) status,
in patients treated with salvage surgery for residual or progressive tumor location, gross tumor volume (GTV), and radiation dose)
disease [13,14]. However, additional surgical resection independent were associated with OS in CR.
of response following CRT was not associated with a survival LRR was classified as recurrence located at the site of the pri-
benefit [14]. mary tumor and/or regional lymph nodes, up to supraclavicular
Patterns and treatment of recurrence in complete responders to nodes. The sites of LRR were assessed in relation to the radiation
definitive CRT for proximal EC are currently unknown, hindering fields and scored as infield or outfield. Distant recurrence was
evidence-based follow-up and clinical counseling of these patients. defined as evidence of disease in any other site. The date of
Whether or not follow-up after definitive CRT for EC is useful re- recurrence was taken as the date of confirmed histology (if present)
mains controversial. Hence, recommendations by NCCN and ESMO or date of imaging of recurrent disease. Diagnostics and treatment
guidelines are inconsistent regarding follow-up strategies and modalities of recurrent disease were assessed.
duration of surveillance [1,3].
The purpose of this study was to determine patterns of recur-
rence, treatment and OS in patients who achieved CR following Statistical analysis
definitive CRT for proximal EC.
Differences between continuous and categorical variables were
tested using the Mann-Witney U and Chi-square test, respectively.
Materials and methods OS was calculated from the start of CRT to date of death. For the
patients with an isolated LRR, OS was additionally calculated from
This multicenter, retrospective, observational study was con- detection of recurrence. PFS was defined as the start of CRT to the
ducted in 11 centers in The Netherlands. Patients were identified in date of recurrence or death. Time to recurrence was calculated from
site-specific databases, complemented with the Netherlands Can- the start of CRT until the occurrence of LRR or distant metastasis,
cer Registry, a population-based cancer registry of all newly diag- whichever came first, with interoccurring death as competing
nosed malignancies in the Netherlands. We identified 200 event. Time to LRR was calculated from the start of CRT to the date
consecutive patients who underwent definitive CRT for proximal of LRR diagnosis, considering interocurring distant metastasis and
EC, from January 2004 to December 2014 [15]. All patients had death as competing events. For all time to event analyses, censoring
histologically confirmed SCC of the proximal esophagus, with occurred at last contact. OS and PFS were calculated using the
maximum distal extension up to 24 cm of the incisors, in which Kaplan-Meier method. Cumulative incidence of recurrence and LRR
supraclavicular nodal involvement was allowed. Tumor staging was were assessed using competing risk analyses. Median follow-up
performed according to the Union for International Cancer Control was calculated using the inverse Kaplan-Meier method for OS
TNM classification that was valid at the time of diagnosis. Since the (death censored). A full multivariable model of potential predictors
TNM classification did not essentially change, no coding was per- for OS was conducted using Cox regression analysis. Multiple
formed to a uniform TNM version, except nodal status which was imputation was used for missing data. Presence of multicollinearity
converted into N1 excluding subdivision of N1/2/3. Further infor- was checked using variation inflation factors.
mation on the study protocol was described previously [15].
Patients were generally examined through physical examination
and history taking in regular follow-up according to national Results
guidelines at four to eight weeks after completion of CRT, and every
three months in the first year, with escalating interval up to five A total of 200 patients were included, of whom 133 patients
years or until death. Endoscopy and imaging was performed in case (67%) achieved a complete response (CR), 42 (21%) a partial
of signs or symptoms of recurrence. No strict imaging protocol for response, and 17 (9%) stable disease following definitive CRT. Five
follow-up was established. CR following CRT was defined as no patients (3%) had progressive disease, and in three patients (2%)
clinical or radiological evidence of locoregional disease three response could not be evaluated.
months after completing CRT as defined by the treating physician. In complete responders, median age at time of diagnosis was 64
Patient informed consent was waived by the Medical Ethics years (range, 42e85 years) (Table 1). Male sex and good perfor-
Board azM/UM due to the retrospective nature of the study (METC mance status, i.e. WHO 0 or 1, were predominant. Median radiation
15-4-012). The study was approved by the scientific committee of dose of the primary treatment was 50.4 Gy. CRT was completed as
the Dutch Upper GI Cancer Group (DUCG), and the Dutch Head and planned in 106 patients (80%). Median follow-up was 64.7 months
Neck Oncology Cooperative Group (NWHHT 2017e01). (95% CI 47.8e81.7).
2017
J. de Vos-Geelen, S.M.E. Geurts, G.A.P. Nieuwenhuijzen et al. European Journal of Surgical Oncology 47 (2021) 2016e2022

Table 1 Table 2
Baseline and tumor characteristics of 133 complete responders following chemo- Multivariable comparison of prognostic factors influencing overall survival in
radiation for proximal esophageal cancer. complete responders (N ¼ 133) following chemoradiation for proximal esophageal
cancer.
Complete
responders HR (95%CI) P value
N ¼ 133
Age at diagnosis 1.00 (0.97-1.04) 0.84
Characteristic No. % Sex
Male Ref.
Age
Female 0.67 (0.39-1.15) 0.15
Median, years (range) 64 (42-85)
WHO performance status
70 years 33 25
0e1 Ref.
Sex
2e3 2.62 (0.90-7.63) 0.08
Male 82 62
Comorbidity
WHO performance score
No Ref.
0-1 118 89
Yes 0.84 (0.50-1.40) 0.50
2-3 6 5
cN stage
Unknown 9 7
N0 Ref.
Pre-diagnostic weight loss
Nþ 1.19 (0.73-1.96) 0.48
<5% 55 41
Tumor location
5e10% 25 19
Cervical Ref.
>10% 26 20
Upper thoracic 0.64 (0.38-1.07) 0.09
Unknown 27 20
GTV (tertiles), cm3
Comorbidity

25 Ref.
Any 75 56
25e43 1.04 (0.56-1.94) 0.90
Cardiovascular disease 40 30
>43 1.64 (0.91-2.97) 0.10
Pulmonary disease 9 7
Radiation dose, Gy
Previous malignancy 23 17

50.4 Ref.
None 58 44
>50.4 1.10 (0.66-1.82) 0.72
Tumor location
Cervical (<18 cm) 41 31 GTV, gross tumor volume: the volume of the macroscopic tumor in cm3 as defined
Upper thoracic (18e24 cm) 92 69 by the tumor-delineation on the radiotherapy planning-CT scan.
Tumor length
Median, cm (range) 4 (1-15)
Obstruction or unknown 38 29 Recurrence
Tumor grade
G1-2 41 31
G3 28 21 Of the 133 patients with a CR, 58 patients had recurrent disease,
Gx 64 48 of whom 32 (55%) had an isolated LRR, 19 (33%) distant metastases
Clinical T stage only, and seven (12%) concurrent locoregional and distant re-
cT1-3 85 64
currences as first site (Fig. 1).
cT4 35 26
cTx 13 10 Three-year PFS was 49% (95% CI 40%e57%). Three- and five-year
Clinical N stage incidences of any recurrence (locoregional or distant) was 40% (95%
cN0 46 35 CI 31%e48%) and 45% (95% CI 36%e54%) and for LRR 26% (95% CI
cNþ 85 64 19e33) and 30% (95% CI 22e38), respectively (Fig. 2).
cNx 2 2
Chemoradiation regimen
Among the patients with LRR, distribution predominantly
Cisplatin, RT > 43.2 and
50.4 Gy 32 24 showed recurrence primarily infield at the original tumor site,
Cisplatin, RT > 50.4 Gy 27 20 followed by infield at the lymph node site (Fig. 1). Metastatic dis-
Carboplatin, Paclitaxel, RT > 43.2 and
50.4 Gy 65 49 ease manifested at a single site in ten patients, and at multiple sites
Carboplatin, Paclitaxel, RT > 50.4 Gy 9 7
in 16 patients. The lungs were the main site of metastatic disease
GTV
Median, cm3 (range) 33.0 (2.4-119.1) (N ¼ 22), followed by distant lymph nodes (N ¼ 9), bone (N ¼ 8),
IQR 19.0e49.1 brain (N ¼ 4), and liver (N ¼ 3).
Unknown 31 23 Of the patients with LRR, 29 patients (74%) were symptomatic at
Radiation dose presentation and in 31 patients (80%) LRR were pathologically
Median, Gy (range) 50.4 (48.6-70.0)
confirmed (Supplementary Table A1). Fourteen patients (54%) who
Chemoradiation completed as planned
Yes 106 80 developed distant metastasis showed physical complaints. Distant
No 21 16 failures were pathologically confirmed in six patients (22%).
Unknown 6 5 Twenty patients developed a second primary tumor, mainly
GTV, gross tumor volume: the volume of the macroscopic tumor in cm3 as defined located in the head and neck (N ¼ 7), lung (N ¼ 4), and esophagus
by the tumor-delineation on the RT planning-CT scan; IQR, interquartile range. (N ¼ 3).
Percentages may not add up to 100 because of rounding. Among the 32 patients with an isolated LRR, 8 (25%) underwent
salvage surgery, 16 (50%) re-irradiation, 3 (9%) chemotherapy, and 8
(25%) best supportive care only. Salvage surgery included esopha-
Overall survival
geal resection and gastric conduit (N ¼ 6), cervical lymph node
dissection (N ¼ 1), and radiofrequency ablation (N ¼ 1). Of the
Median OS was 45.0 months (95% CI 34.8e61.5 months) and
patients undergoing surgery for their recurrence, four received
three-year OS was 58% (95% CI 48%e66%) (Supplementary
multimodal treatment with RT (N ¼ 3) or CRT (N ¼ 1). Median
Figure A1). Neither age, sex, comorbidity, lymph node status,
radiation dose to treat LRR was 30 Gy (range 12e50.4 Gy).
GTV, nor radiation dose were significant prognostic factors for OS,
Median OS of patients undergoing salvage surgery for isolated
whereas WHO performance status and tumor location were iden-
LRR was 50.6 months (95% CI 31.2e70.0, three- and five-year OS
tified as clinically important factors with borderline significance
75% and 30%) from the start of CRT at primary diagnosis. Median OS
(Table 2).
from detection of LRR was 32.0 months (95% CI 6.8-not reached,

2018
J. de Vos-Geelen, S.M.E. Geurts, G.A.P. Nieuwenhuijzen et al. European Journal of Surgical Oncology 47 (2021) 2016e2022

Fig. 1. Location of first recurrence in complete responders following chemoradiation for proximal esophageal cancer.
LRR, locoregional recurrence.

three- and five-year OS 44% and 29%) for salvage surgery, 8.7 one out of five patients with either a locoregional or distant
months for patients undergoing re-irradiation with or without recurrence received chemotherapy.
chemotherapy, 3.0 months for chemotherapy alone, and 1.5 months To the best of our knowledge, this is the largest cohort reporting
for best supportive care only (Supplementary Figure A2). At the end the pattern of recurrence in complete responders after CRT for
of follow-up, five patients had died following salvage surgery, all as proximal EC, observing a three- and five-year risk of recurrence of
a result of recurrent disease. Timing of detection of LRR in patients 40% and 45%. Two-thirds of all recurrences were locoregional fail-
undergoing salvage surgery compared with those patients with LRR ures, which was in line with previous cohorts in proximal EC,
only who were unable to undergo surgery for LRR did not differ reporting LRR in 58e84% of the recurrences after CRT although not
(Supplementary Figure A3). specific for CR [4,5,8,14,16,17], except for a small Canadian cohort of
In metastatic disease (with or without LRR), three patients (12%) complete responders (N ¼ 38) [4].
underwent a metastasectomy (with or without RT). Their median The high rate of LRR after CRT emphasizes consideration of
OS from detection of recurrence was 16.0 months. Four patients improvement in local therapy. However, adjuvant resection after
(15%) underwent chemotherapy (with or without RT) with a me- CRT did not prolong survival in cervical EC compared with defini-
dian OS of 7.0 months. A variety of chemotherapeutic agents were tive CRT alone in an Italian cohort study [14]. Nonetheless, ran-
administered in recurrent disease, mostly fluoropyrimidine-based. domized data are lacking.
RT only was applied in five patients with a median radiation dose of Locoregional recurrences were mainly within the RT field, sug-
22.8 Gy (range 18e39 Gy), demonstrating a median OS of 7.6 gesting a potential benefit of escalating radiation dose. However,
months. Median OS for patients not receiving antitumor treatment previous studies showed inconsistent results of dose escalation. A
(N ¼ 14) was 3.5 months. recent meta-analysis including ten observational studies
(N ¼ 4918), and one small prospective trial (N ¼ 28), suggested that
Discussion a higher radiation dose to the primary tumor volume was associ-
ated with better locoregional failure free survival [18]. However, the
This retrospective cohort study demonstrated that most re- recent phase III dose escalation ARTDECO study in EC randomizing
currences in complete responders following CRT for proximal EC 260 patients (61% SCC), demonstrated that radiation dose escala-
were locoregional and occurred mainly in the primary involved tion up to 61.6 Gy versus 50.4 Gy to the primary tumor did not
tumor field within the first three years. In most cases of LRR local improve local control [19].
therapy was applied, i.e. salvage surgery or second course RT. Only Salvage surgery in case of an isolated LRR could be considered

2019
J. de Vos-Geelen, S.M.E. Geurts, G.A.P. Nieuwenhuijzen et al. European Journal of Surgical Oncology 47 (2021) 2016e2022

(38%) with LRR of cervical EC who underwent surgery, of which

three remained alive for more than three years [13]. Schieman et al.
demonstrated corresponding outcomes, with a three-year OS rate
of 33% [20]. The reported long-term survival following salvage
surgery was confirmed in retrospective studies including patients
with cancers of the esophagus, i.e. proximal, mid or distal, treated
with definitive CRT and experiencing a LRR [21,22]. Moreover, re-
currences in the proximal part of the esophagus are even more
difficult to treat due to long-term side effects after definitive CRT,
e.g. strictures and fistulas [23e25]. Hence, most patients suffering
from a LRR following CRT for proximal EC are unable to undergo
salvage surgery. Although a durable survival can be achieved in this
group, salvage esophagectomy should be thoughtfully balanced
with perioperative morbidity and mortality of such challenging
resections by both patients and physicians [21,22].
The role of re-irradiation in LRR following definitive CRT in EC is
controversial. In the current study, re-irradiation was carried out in
16 patients with an isolated LRR, with a median radiation dose of
30 Gy (range 12e50.4 Gy). Median OS of patients undergoing
salvage RT was 8.7 months from the point of detection of locore-
gional disease. Literature on re-irradiation in EC is sparse. Zhou
et al. retrospectively analyzed 55 patients with recurrences treated
with salvage RT with a median dose of 54 Gy (range 18e66 Gy),
demonstrating a median OS of five months [26]. A recent Chinese
study established long-term survival of 17 months in patients
treated with re-irradiation for locoregional recurrent esophageal
SCC [27]. Others found comparable poor outcomes following re-
irradiation [28,29]. Hence, only in selected and highly motivated
patients a radical radiotherapeutic approach can be considered,
outweighing the significant risks of re-irradiation, e.g. fistulation,
stenosis, and vascular blow-out.
Although the minority of initial recurrences in our study were
distant, improvements in systemic therapy remain warranted,
regarding optimization of systemic cytotoxic effects in recurrent
disease, but also for synergistic effects in primary CRT. The imple-
mentation of immunotherapy in metastatic EC is to be awaited,
after proved safety and efficacy of the ATTRACTION-3, CheckMate-
649, and KEYNOTE-590 trials, especially in the subset of pro-
grammed death ligand-1 positive tumors [30e32]. Furthermore,
the adjuvant CheckMate-577 trial showed a significantly improved
disease-free survival with adjuvant nivolumab compared with
placebo in patients with resected EC after neoadjuvant chemo-
radiation and have not achieved a pathological complete response
[33]. In addition, a phase II study is assessing the efficacy of ate-
zolizumab following definitive CRT to increase CR rate [34].
In the current study, main metastatic sites were lung and lymph
nodes, comparable with the results from smaller observational
studies [8,35,36]. In our cohort, only 15% of patients developing
distant metastases received palliative chemotherapy. This corre-
sponds to historical data in a large cohort of proximal EC [37], and
reflects the limited high level evidence of palliative systemic ther-
apy in metastatic esophageal SCC [38], as well as the frailty of this
population.
The approach regarding evaluation of response after CRT has not
been established. ESMO guidelines do not include recommenda-
Fig. 2. Progression-free survival (A), cumulative incidence of recurrence (locoregional
or distant) (B), and cumulative incidence of locoregional recurrence as first event (C) tions concerning response evaluation following definitive CRT [3].
for patients with complete response following chemoradiation for proximal esopha- However, the NCCN guideline committee recommends endoscopy
geal cancer. and biopsy [1]. Others have suggested a role for 18F-FDG PET/CT,
Dashed lines correspond with the 95% confidence intervals. diffusion-weighted MRI [39] or molecular biomarkers, such as
circulating tumor DNA [40] for response evaluation, which has to
be explored in future studies. The potential benefits of such
for highly selected patients. In our study eight out of 32 patients
extensive assessments have to outweigh the impact on either pa-
(25%) with LRR only underwent salvage surgery, of which 44% were
tients and healthcare facilities.
alive at three years after detection of recurrence. These results are
Systematic surveillance strategies after successful definitive
comparable with a Japanese phase II trial including five patients
treatment for proximal EC remain controversial. We demonstrated
2020
J. de Vos-Geelen, S.M.E. Geurts, G.A.P. Nieuwenhuijzen et al. European Journal of Surgical Oncology 47 (2021) 2016e2022

that most patients presented with symptoms at the time of recur- CRediT authorship contribution statement
rence, which was expected since Dutch esophageal and head and
neck cancer guidelines, as well as NCCN guidelines [1], advice J. de Vos-Geelen: Conceptualization, Methodology, Software,
symptom-based follow-up, whereas endoscopy and imaging Validation, Investigation, Writing - original draft, Project adminis-
studies are only recommended to be performed on indication. In tration. S.M.E. Geurts: Conceptualization, Methodology, Software,
contrast, ESMO guideline states that a three-month follow-up Validation, Formal analysis, Investigation, Writing - review &
based on endoscopy, biopsies and CT scan may be recommended in editing. G.A.P. Nieuwenhuijzen: Conceptualization, Resources,
the case of CR to definitive CRT. It might be expected that a more Writing - review & editing. F.E.M. Voncken: Resources, Writing -
vigorous approach of follow-up will lead to an earlier detection of review & editing. J.A. Bogers: Resources, Writing - review & edit-
recurrences. An earlier detection may, in addition, lead to an ing. P.M. Braam: Resources, Writing - review & editing. C.T. Muijs:
increased rate of effective salvage interventions. However, associ- Resources, Writing - review & editing. M.A. de Jong: Resources,
ated outcome remains unknown. The usefulness of follow-up in Writing - review & editing. N. Kasperts: Resources, Writing - re-
terms of improving survival is limited for early salvage surgery after view & editing. T. Rozema: Resources, Writing - review & editing.
(failing) definitive CRT [3]. It would be of great interest to study the G.J. Blom: Resources, Writing - review & editing. S.A.W. Bouwense:
influence of the different methods of follow-up on the rate of Writing - review & editing, Visualization, Writing - review &
detection of LRR, and the chances of performing salvage treatment. editing. L.B.J. Valkenburg-van Iersel: Resources, Writing - review
Particularly in the current era of enhanced surgical procedures in & editing. P.M. Jeene: Resources, Writing - review & editing. F.J.P.
the salvage setting for EC, e.g. endoscopic resections [41], robot- Hoebers: Conceptualization, Methodology, Writing - review &
assisted minimally invasive esophagectomy [42], and the advan- editing, Supervision. V.C.G. Tjan-Heijnen: Conceptualization,
tages of high volume expertise centers [43,44]. Methodology, Writing - review & editing, Supervision, Funding
In the Netherlands, follow-up ends five years post treatment. acquisition.
Duration of follow-up is not clearly specified in ESMO guidelines,
whereas NCCN recommends annual follow-up even after five years. Declaration of competing interest
The current study demonstrated that most failures were developed
within the first three years after CRT. After that point, surveillance JV has served as a consultant for AstraZeneca, MSD, Pierre Fabre,
for second primary tumors may be of importance, considering the and Servier, and has received institutional research funding from
high occurrence of a second SCC due to the close association of Servier. All outside the submitted work.
alcohol consumption and smoking habits [45]. Whether early SG has received institutional research funding from Roche,
detection and treatment of second primaries improves patient Pfizer, Novartis, and Eli Lilly. All outside the submitted work.
outcomes is currently unknown. Considering the impact of pro- CM had research collaborations with IBA, Siemens, Raystation,
longed follow-up might have on patients quality of life and health and Mirada. All outside the submitted work.
care costs, we would propose to restrict follow-up to three years VT has received honoraria/travel grants from Roche, Novartis,
following CR after definitive CRT for proximal EC. Unless other Pfizer, Lilly, and Accord Healthcare, and has received institutional
factors regarding patients recovery, e.g. repeated dilatations, di- research funding from AstraZeneca, Roche, Pfizer, Novartis, Eisai,
etary or psychosocial needs, require continued surveillance. and Lilly. All outside the submitted work.
Furthermore, follow-up should be patient tailored in order to The other authors have declared no conflicts of interest.
warrant patients’ preferences.
The strength of our study is that we included a large cohort of Appendix A. Supplementary data
patients in this rare disease, with long-term follow-up. The retro-
spective design of this study is however also inherent with some Supplementary data to this article can be found online at
limitations. Details on applied RT techniques, time to response, https://doi.org/10.1016/j.ejso.2021.02.001.
subsequent recurrences were not collected. Furthermore, the
number of patients with recurrences did not allow us to study in-
References
dependent predictive and prognostic factors for risk of LRR and
radiation dose effect on treatment response. In addition, recurrence [1] Ajani JA, D'Amico TA, Bentrem DJ, Chao J, Corvera C, Das P, et al. Esophageal
treatment outcomes should be considered with care, given the low and esophagogastric junction cancers, version 2.2019, NCCN clinical practice
patient numbers. guidelines in oncology. J Natl Compr Canc Netw 2019;17(7):855e83.
[2] Shah Manish A, Kennedy Erin B, Catenacci Daniel V, Deighton Dana C,
In conclusion, patients with a complete response after definitive Goodman Karyn A, Malhotra Narinder K, et al. Treatment of locally advanced
CRT for proximal EC demonstrated high rates of infield LRR, sug- esophageal carcinoma: ASCO guideline. J Clin Oncol 2020;38(23):2677e94.
gesting methods to optimize locoregional control are necessary. [3] Lordick F, Mariette C, Haustermans K, Obermannova R, Arnold D. Oesophageal
cancer: ESMO Clinical Practice Guidelines for diagnosis, treatment and follow-
Salvage treatment of isolated LRR resulted in favorable outcome. up. Ann Oncol 2016;27(suppl 5):v50e7.
Hence, screening for LRR should be optimized. Furthermore, since [4] Huang SH, Lockwood G, Brierley J, Cummings B, Kim J, Wong R, et al. Effect of
nearly all recurrences occurred within three years after initial concurrent high-dose cisplatin chemotherapy and conformal radiotherapy on
cervical esophageal cancer survival. Int J Radiat Oncol Biol Phys 2008;71(3):
treatment, routine follow-up could be restricted. 735e40.
[5] Zhang P, Xi M, Zhao L, Qiu B, Liu H, Hu YH, et al. Clinical efficacy and failure
pattern in patients with cervical esophageal cancer treated with definitive
chemoradiotherapy. Radiother Oncol 2015;116(2):257e61.
Funding [6] Herrmann E, Mertineit N, De Bari B, Hoeng L, Caparotti F, Leiser D, et al.
Outcome of proximal esophageal cancer after definitive combined chemo-
None. radiation: a Swiss multicenter retrospective study. Radiat Oncol 2017;12(1):
97.
[7] Gkika E, Gauler T, Eberhardt W, Stahl M, Stuschke M, Pottgen C. Long-term
results of definitive radiochemotherapy in locally advanced cancers of the
Data statement cervical esophagus. Dis Esophagus 2014;27(7):678e84.
[8] Yamada K, Murakami M, Okamoto Y, Okuno Y, Nakajima T, Kusumi F, et al.
Treatment results of radiotherapy for carcinoma of the cervical esophagus.
The data that support the findings of this study are available Acta Oncol 2006;45(8):1120e5.
from the corresponding author upon request. [9] Uno T, Isobe K, Kawakami H, Ueno N, Shimada H, Matsubara H, et al.

2021
J. de Vos-Geelen, S.M.E. Geurts, G.A.P. Nieuwenhuijzen et al. European Journal of Surgical Oncology 47 (2021) 2016e2022

Concurrent chemoradiation for patients with squamous cell carcinoma of the after primary definitive radiotherapy. Radiat Oncol J 2012;30(4):182e8.
cervical esophagus. Dis Esophagus 2007;20(1):12e8. [29] Yamaguchi Shinsaku, Ohguri Takayuki, Imada Hajime, Yahara Katsuya,
[10] Ludmir EB, Palta M, Zhang X, Wu Y, Willett CG, Czito BG. Incidence and Moon Dae Seung, Higure Aiichiro, et al. Multimodal approaches including
prognostic impact of high-risk HPV tumor infection in cervical esophageal three-dimensional conformal Re-irradiation for recurrent or persistent
carcinoma. J Gastrointest Oncol 2014;5(6):401e7. esophageal cancer: preliminary results. J Radiat Res 2011;52(6):812e20.
[11] Hoeben A, Polak J, Van De Voorde L, Hoebers F, Grabsch HI, de Vos-Geelen J. [30] Kato K, Cho BC, Takahashi M, Okada M, Lin CY, Chin K, et al. Nivolumab versus
Cervical esophageal cancer: a gap in cancer knowledge. Ann Oncol chemotherapy in patients with advanced oesophageal squamous cell carci-
2016;27(9):1664e74. noma refractory or intolerant to previous chemotherapy (ATTRACTION-3): a
[12] de Vos-Geelen J, Hoebers FJP, Geurts SME, Hoeben A, de Greef BTA, multicentre, randomised, open-label, phase 3 trial. Lancet Oncol 2019;20(11):
Voncken FEM, et al. A national study to assess outcomes of definitive che- 1506e17.
moradiation regimens in proximal esophageal cancer. Acta Oncol 2020;59(8): [31] Moehler M. Nivolumab (NIVO) plus chemotherapy (chemo) versus chemo as
895e903. first-line (1l) treatment for advanced gastric cancer/gastroesophageal junction
[13] Zenda S, Kojima T, Kato K, Izumi S, Ozawa T, Kiyota N, et al. Multicenter phase cancer (GC/GEJC)/esophageal adenocarcinoma (EAC): first results of the
2 study of cisplatin and 5-fluorouracil with concurrent radiation therapy as an CheckMate 649 study. Ann Oncol 2020;31. Abstract LBA6_PR.
organ preservation approach in patients with squamous cell carcinoma of the [32] Kato K. Pembrolizumab plus chemotherapy versus chemotherapy as first-line
cervical esophagus. Int J Radiat Oncol Biol Phys 2016;96(5):976e84. therapy in patients with advanced esophageal cancer: the phase 3 KEYNOTE-
[14] Valmasoni M, Pierobon ES, Zanchettin G, Briscolini D, Moletta L, Ruol A, et al. 590 study. Ann Oncol 2020;31. Abstract LBA8_PR.
Cervical esophageal cancer treatment strategies: a cohort study appraising the [33] Kelly RJ. Adjuvant nivolumab in resected esophageal or gastroesophageal
debated role of surgery. Ann Surg Oncol 2018;25(9):2747e55. junction cancer (EC/GEJC) following neoadjuvant chemoradiation therapy
[15] de Vos-Geelen J, Hoebers FJP, Geurts SME, Hoeben A, de Greef BTA, (CRT): first results of the CheckMate 577 study. Ann Oncol 2020;31:
Voncken FEM, et al. A national study to assess outcomes of definitive che- S1142e215.
moradiation regimens in proximal esophageal cancer. Acta Oncol 2020;59(8): [34] Bando H, Kotani D, Tsushima T, Hara H, Kadowaki S, Kato K, et al. TENERGY:
895e903. multicenter phase II study of Atezolizumab monotherapy following definitive
[16] Zhao L, Zhou Y, Mu Y, Chai G, Xiao F, Tan L, et al. Patterns of failure and clinical Chemoradiotherapy with 5-FU plus Cisplatin in patients with unresectable
outcomes of definitive radiotherapy for cervical esophageal cancer. Onco- locally advanced esophageal squamous cell carcinoma. BMC Canc 2020;20(1):
target 2017;8(13):21852e60. 336.
[17] Esmati E, Maddah Safaei A, Ghalehtaki R, Mousavi N, Saraee E, Shirouei S, et al. [35] Cao CN, Luo JW, Gao L, Xu GZ, Yi JL, Huang XD, et al. Primary radiotherapy
Outcomes of definitive chemoradiotherapy for cervical and upper thoracic compared with primary surgery in cervical esophageal cancer. JAMA Otolar-
esophageal cancers: a single-institution experience of a rare cancer. yngol Head Neck Surg 2014;140(10):918e26.
J Gastrointest Canc 2019;50(3):380e5. [36] Cao C, Luo J, Gao L, Xu G, Yi J, Huang X, et al. Definitive radiotherapy for
[18] Xiao L, Czito BG, Pang Q, Hui Z, Jing S, Shan B, et al. Do higher radiation doses cervical esophageal cancer. Head Neck 2015;37(2):151e5.
with concurrent chemotherapy in the definitive treatment of esophageal [37] de Vos-Geelen J, Geurts SM, van Putten M, Valkenburg-van Iersel LB,
cancer improve outcomes? A meta-analysis and systematic review. J Canc Grabsch HI, Haj Mohammad N, et al. Trends in treatment and overall survival
2020;11(15):4605e13. among patients with proximal esophageal cancer. World J Gastroenterol
[19] Hulshof MCCM, Geijsen D, Rozema T, Oppedijk V, Buijsen J, Neelis KJ, et al. 2019;25(47):6835e46.
A randomized controlled phase III multicenter study on dose escalation in [38] Janmaat Vincent T, Steyerberg Ewout W, van der Gaast Ate, Mathijssen Ron
definitive chemoradiation for patients with locally advanced esophageal HJ, Bruno Marco J, Peppelenbosch Maikel P, et al. Palliative chemotherapy and
cancer: ARTDECO study. J Clin Oncol 2020;38. Abstract 281. targeted therapies for esophageal and gastroesophageal junction cancer.
[20] Schieman C, Wigle DA, Deschamps C, Nichols 3rd FC, Cassivi SD, Shen KR, et al. Cochrane Database Syst Rev 2017;11(11):CD004063.
Salvage resections for recurrent or persistent cancer of the proximal esoph- [39] Borggreve AS, Goense L, van Rossum PSN, Heethuis SE, van Hillegersberg R,
agus after chemoradiotherapy. Ann Thorac Surg 2013;95(2):459e63. Lagendijk JJW, et al. Preoperative prediction of pathologic response to neo-
[21] Markar Sheraz, Gronnier Caroline, Duhamel Alain, Pasquer Arnaud, adjuvant chemoradiotherapy in patients with esophageal cancer using (18)F-
The
reaux Je
re

mie, Rieu Mael Chalret du, et al. Salvage surgery after chemo- FDG PET/CT and DW-MRI: a prospective multicenter study. Int J Radiat Oncol
radiotherapy in the management of esophageal cancer: is it a viable thera- Biol Phys 2020;106(5):998e1009.
peutic option? J Clin Oncol 2015;33(33):3866e73. [40] Azad TD, Chaudhuri AA, Fang P, Qiao Y, Esfahani MS, Chabon JJ, et al. Circu-
[22] Sudo K, Xiao L, Wadhwa R, Shiozaki H, Elimova E, Taketa T, et al. Importance lating tumor DNA analysis for detection of minimal residual disease after
of surveillance and success of salvage strategies after definitive chemo- chemoradiotherapy for localized esophageal cancer. Gastroenterology
radiation in patients with esophageal cancer. J Clin Oncol 2014;32(30): 2020;158(3):494e505. e496.
3400e5. [41] Al-Kaabi A, Schoon EJ, Deprez PH, Seewald S, Groth S, Giovannini M, et al.
[23] Faiz Z, Dijksterhuis WPM, Burgerhof JGM, Muijs CT, Mul VEM, Wijnhoven BPL, Salvage endoscopic resection after definitive chemoradiotherapy for esoph-
et al. A meta-analysis on salvage surgery as a potentially curative procedure in ageal cancer: a Western experience. Gastrointest Endosc
patients with isolated local recurrent or persistent esophageal cancer after 2020;S0016e5107(20). 34657-5.
chemoradiotherapy. Eur J Surg Oncol 2019;45(6):931e40. [42] Defize IL, van der Horst S, Ruurda JP, van Hillegersberg R. ASO author re-
[24] Wang S, Liao Z, Chen Y, Chang JY, Jeter M, Guerrero T, et al. Esophageal cancer flections: preoperative selection of cT4b esophageal cancer patients who
located at the neck and upper thorax treated with concurrent chemo- benefit from a salvage robot-assisted minimally invasive esophagectomy
radiation: a single-institution experience. J Thorac Oncol 2006;1(3):252e9. (RAMIE). Ann Surg Oncol 2020.
[25] Tu L, Sun L, Xu Y, Wang Y, Zhou L, Liu Y, et al. Paclitaxel and cisplatin com- [43] van Putten M, Koe €ter M, van Laarhoven HWM, Lemmens Vepp, Siersema PD,
bined with intensity-modulated radiotherapy for upper esophageal carci- Hulshof MCCM, et al. Hospital of diagnosis influences the probability of
noma. Radiat Oncol 2013;8:75. receiving curative treatment for esophageal cancer. Ann Surg 2018;267(2):
[26] Zhou Zhi-guo, Zhen Chan-jun, Bai Wen-wen, Zhang Ping, Qiao Xue-ying, 303e10.
Liang Jun-li, et al. Salvage radiotherapy in patients with local recurrent [44] van de Poll-Franse LV, Lemmens VE, Roukema JA, Coebergh JW,
esophageal cancer after radical radiochemotherapy. Radiat Oncol 2015;10(1): Nieuwenhuijzen GA. Impact of concentration of oesophageal and gastric
54. cardia cancer surgery on long-term population-based survival. Br J Surg
[27] Xu X, Wang Z, Jiang S, Shang Y, Wu Y. Evaluating the optimal re-irradiation 2011;98(7):956e63.
dose for locally recurrent esophageal squamous cell carcinoma after defini- [45] Mitani S, Kadowaki S, Oze I, Masuishi T, Narita Y, Bando H, et al. Risk of second
tive radiotherapy. Radiat Oncol 2019;14(1):191. primary malignancies after definitive treatment for esophageal cancer: a
[28] Kim Young Suk, Lee Chang Geol, Kim Kyung Hwan, Kim Taehyung, competing risk analysis. Canc Med 2020;9(1):394e400.
Lee Joohwan, Cho Yona, et al. Re-irradiation of recurrent esophageal cancer

2022