BJR https://doi.org/10.1259/bjr.

20220410
© 2022 The Authors. Published by the British Institute of Radiology under
Received: Accepted: Published online: the terms of the Creative Commons Attribution 4.0 Unported License
10 April 2022 06 June 2022 15 June 2022 http://creativecommons.org/licenses/by/4.0/, which permits unrestricted
use, distribution and reproduction in any medium, provided the original
author and source are credited.

Cite this article as:

Withey SJ, Goh V, Foley KG. State-­of-­the-­art imaging in oesophago-­gastric cancer. Br J Radiol (2022) 10.1259/bjr.20220410.

REVIEW ARTICLE

State-­of-­the-­art imaging in oesophago-­gastric cancer

1
SAMUEL J WITHEY, FRCR, 2,3VICKY GOH, MD, FRCR and 4,5KIERAN G FOLEY, PhD, FRCR
1
Department of Radiology, The Royal Marsden NHS Foundation Trust, London, UK
2
Cancer Imaging, School of Biomedical Engineering & Imaging Sciences, King’s College London, London, UK
3
Department of Radiology, Guy’s and St Thomas’ NHS Foundation Trust, London, UK
4
Division of Cancer & Genetics, School of Medicine, Cardiff University, Wales, UK
5
Department of Radiology, Velindre Cancer Centre, Cardiff, UK

Address correspondence to: Dr Kieran G Foley

E-mail: [email protected]

ABSTRACT
Radiological investigations are essential in the management of oesophageal and gastro-­oesophageal junction cancers.
The current multimodal combination of CT, 18F-­fluorodeoxyglucose positron emission tomography combined with CT
(PET/CT) and endoscopic ultrasound (EUS) has limitations, which hinders the prognostic and predictive information
that can be used to guide optimum treatment decisions. Therefore, the development of improved imaging techniques
is vital to improve patient management. This review describes the current evidence for state-­of-­the-­art imaging tech-
niques in oesophago-­gastric cancer including high resolution MRI, diffusion-­weighted MRI, dynamic contrast-­enhanced
MRI, whole-­body MRI, perfusion CT, novel PET tracers, and integrated PET/MRI. These novel imaging techniques may
help clinicians improve the diagnosis, staging, treatment planning, and response assessment of oesophago-­gastric
cancer.

INTRODUCTION TNM classification, but the overall stage groups have subtle
Around 9000 cases of oesophageal and gastro-­oesophageal differences between subtypes.6
junction (GOJ) cancers (referred to collectively in this
review as oesophago-­gastric cancers) are diagnosed in the In terms of management, tumours confined to the mucosa
UK each year and the incidence has risen by 6% over the (T1a) can be considered for endoscopic resection or abla-
past 20 years.1 Worldwide, there are more than 600,000 tion, whereas tumours involving the submucosa (T1b)
new cases each year.2 Squamous cell carcinoma (SCC) usually require oesophagectomy due to higher rates of
is the most common subtype worldwide but adenocar- occult lymph node metastases.8 Neoadjuvant therapy
cinoma is more common in the UK, USA, and Western provides a survival benefit for patients with locally advanced
Europe.3 Both are associated with poor prognosis, with disease but the benefit reduces in early tumours.9 Regional
nodal involvement is a further indication for neoadjuvant
5-­year survival reported at 12% for SCC and 15% for
therapy.9 Distant metastatic disease is present in up to 50%
adenocarcinoma.4
of patients at diagnosis, and generally precludes surgical
management,10 although the value of surgical and ablative
SCCs are distributed equally between the upper and mid-­
techniques in oligometastatic disease is being explored.11
thoracic oesophagus, whereas most adenocarcinomas are
located in the distal oesophagus or at the GOJ.5 Using the
Imaging is essential for all aspects of oesophago-­gastric
current TNM eighth edition classification, Siewert Type cancer management. In this review, we summarise current
1 and 2 GOJ tumours are staged as oesophageal cancers, best practice and highlight how state-­of-­the-­art imaging
and Siewert Type 3 tumours (with the epicentre in the can improve diagnosis and staging, allow more effective
proximal stomach between 2 and 5 cm from the GOJ) treatment planning and monitoring, and improve risk
are staged as gastric cancers.6 The oesophagus has a rich stratification.
bidirectional lymphatic drainage system, meaning lymph
node metastases (and satellite nodules) can develop along Current best practice in diagnosis and staging
the entire length of the oesophagus.7 Histological subtype Oesophago-­gastric cancer is usually diagnosed following
is generally not taken into account when assigning the first-­
line endoscopy and biopsy, but disease staging is
 BJR Withey et al

Figure 1. Typical radiological staging pathway for most patients diagnosed with oesophago-­gastric cancer in the United Kingdom.
EUS, endoscopic ultrasound; PET, positron emission tomography.

largely influenced by radiological techniques. A patient staging the oesophageal wall, therefore is inaccurate for early T-­staging.14
algorithm describing the current radiological pathway is in When dichotimising early (T1-­T2) vs late (T3-­T4) oesophageal
shown in Figure  1. The TNM v. 8.0 staging classification for cancer, CT has a reported diagnostic accuracy of 80–82% (n = 74
oesophago-­gastric cancer is described in Table 1. patients).15 Using a 1 cm short-­axis size threshold, the sensitivity
in diagnosing lymph node metastases has been reported as 50%,
CT and specificity as 83%.16
Contrast-­ enhanced CT is the primary radiological staging
investigation (Figure 2), usually performed after endoscopy and Endoscopic ultrasound
biopsy.12,13 CT can identify patients with clear distant metastatic Endoscopic ultrasound (EUS) is used in some centres for more
disease which usually signals palliative therapy and precludes detailed locoregional staging (Figure 2).16 EUS may be used to
futile radical treatment. However, CT cannot identify the layers of differentiate between T1a and T1b tumours17 although when

Table 1. TNM staging system for oesophageal carcinomas

T Category T Criteria
TX Tumour cannot be assessed
T0 No evidence of primary tumour
Tis High-­grade dysplasia, defined as malignant cells confined to the epithelium by the basement membrane
T1a Tumour invades the lamina propria or muscularis mucosae
T1b Tumour invades the submucosa
T2 Tumour invades the muscularis propria
T3 Tumour invades adventitia
T4a Tumour invades the pleura, pericardium, azygos vein, diaphragm, or peritoneum
T4b Tumour invades other adjacent structures, such as the aorta, vertebral body, or airway
N Category N Criteria
NX Regional lymph nodes cannot be assessed
N0 No regional lymph node metastasis
N1 Metastases in one or two regional lymph nodes
N2 Metastases in three to six regional lymph nodes
N3 Metastases in seven or more regional lymph nodes
M Category M Criteria
M0 No distant metastases
M1 Distant metastases
Reproduced from: AJCC Cancer Staging Manual. eighth Edition ed: Springer International Publishing, 2017.

2 of 12 birpublications.org/bjr Br J Radiol;95:20220410
 State-­of-­the-­art imaging in oesophago-­gastric cancer BJR

Figure 2. Axial CT showing mural thickening of the distal oesophagus (a), axial fused 18F-­FDG PET/CT (b) showing FDG uptake in
the tumour. Clinical staging of both CT and PET/CT was cT3 N0 M0. Endoscopic ultrasound (c) showed that the distal oesophageal
tumour (arrow) involved the diaphragmatic crus distally and there was a malignant perioesophageal lymph node (arrowhead),
therefore the final staging was cT4a N1 M0 (Images courtesy of Dr K G Foley, Velindre Cancer Centre). FDG, fluorodeoxyglucose;
PET, positron emission tomography.

dichotimising T1-­T2 vs T3-­T4 oesophageal cancer, EUS has a potentially resectable oesophageal cancer develop metastases
reported diagnostic accuracy of 81–85% (n = 74 patients).15 Its whilst on neoadjuvant chemotherapy.22,23 However, complete
sensitivity and specificity for nodal metastases is 80 and 70%, pathological response is reported in 32–52% depending on the
respectively.16 A benefit of EUS is that it provides an oppor- criteria used24–26 and a management aspiration is that some of
tunity to perform fine needle aspiration (FNA) of suspicious these patients may be candidates for surveillance programmes
nodes, increasing its diagnostic accuracy from 74 to 87%.18 The rather than oesophagectomy in future.
limitations of EUS include its availability, accuracy related to
operator experience, and stenotic tumours not traversable by the Contrast-­enhanced CT has low sensitivity for residual disease,
endoscope.19 so cannot adequately assess for treatment response.27 EUS is
also of little value in restaging after neoadjuvant therapy because
PET/CT post-­treatment inflammation and fibrosis can be indistin-
18-­Fluorine fluorodeoxyglucose (18F-­FDG) PET/CT is recom- guishable from residual tumour.28 A further challenge is that
mended for assessment of metastatic disease undetected by nodal response can be discordant with the primary tumour (in
CT in those planned for curative treatment. 18F-­FDG PET/CT approximately 5%), and prognosis may be improved in cases
has a greater sensitivity for distant metastases than CT (71% with improved nodal disease, even in the absence of response
vs  52%).16 However, the sensitivity for perioesophageal nodal in the primary tumour.29 Conventional morphological imaging
disease is poor, reported at 57%, with 85% specificity.16 This is, may struggle to assess this, as even morphologically normal
in part, due to the spatial resolution of PET limiting the differen- lymph nodes can contain metastasis. Foley et al reviewed resec-
tiation of perioesophageal nodes from the primary tumour. Use tion specimens in 15 patients pre-­operatively staged as N0 but
of 18F-­FDG PET/CT has been reported to improve patient strat- with nodal metastases found on pathological assessment. In 50
ification, reduce relapse rate, and increase overall survival after nodal metastases, 22% were 2 mm or less, and 82% were 6 mm or
oesophagectomy.20 In surgical candidates, clinically relevant less, implying that novel methods that augment morphological
changes to staging have been reported in around 24%, mainly assessment are required.30
related to upstaging to M1 disease.21
CURRENT BEST PRACTICE IN FOLLOW-UP &
CURRENT BEST PRACTICE IN IMAGING TO SUSPECTED RECURRENCE
ASSESS RESPONSE Following treatment, the final challenge for imaging is to monitor
CT is routinely performed, and 18F-­ FDG PET/CT in some for disease relapse. Despite multimodality therapy, there are high
centres, after neoadjuvant therapy to assess response, although rates of post-­treatment relapse, reported at 45–53% within 2
in current practice, this is simply to ensure that disease has not years of surgery.31,32 Anastomotic recurrence following surgery
progressed and become unresectable (either due to advanced occurs in 7–12%, seen as nodular or concentric thickening in the
T-­stage or development of metastases). Up to 10% of patients with region of the anastomosis on CT.31,33 In one study of recurrence

3 of 12 birpublications.org/bjr Br J Radiol;95:20220410
 BJR Withey et al

Figure 3. Small field-­of-­view axial T2 weighted MRI. There is MRI with diffusion-­weighted sequences (DWI) has the potential
a primary tumour centred on the left-­side of the oesophagus to improve staging accuracy and assessment of tumour length.38
extending through the muscularis into the perioesophageal A recent systematic review identified 984 patients in 19 studies
fat (arrow). The normal intact layers of the oesophageal wall and found that MRI (without any restriction on the sequences
are seen on the contralateral side (Image courtesy of Dr A M used) had a sensitivity of 67–91% and specificity of 91–92% for
Riddell, Royal Marsden Hospital). differentiating T2 tumours or less from T3 or above, compared
to 85 and 75% for CT, and 68–100% and 75–100% for EUS.38–40

Notably, oesophageal MRI is challenged by organ peristalsis,

cardiac and respiratory motion, aortic blood flow and pulsation,
and artefact from medical devices. Additionally, due to its central
position within the thorax, the oesophagus is distant from MR
coils which lowers the signal-­to-­noise ratio.41 Use of 3 T MRI
systems can increase the signal-­to-­noise ratio but at the cost of
increased susceptibility artefact. Cardiac and respiratory gating
and endoluminal or surface coils can improve image quality.36,42

In terms of differentiating node negative from node positive

disease, MRI has a reported sensitivity of 59–100% and speci-
ficity 57–92%, dependent on the size threshold used to define a
metastatic lymph node,39 compared to 83 and 75% for CT, 46 and
91% for PET/CT, and 100 and 36% for EUS.38,39,43 In addition
to size criteria, alternative imaging features have been proposed
following oesophagectomy, 50% of the 435 recurrences were to differentiate malignant vs normal nodes. Alper et al used
detected as a result of symptoms, and 45% were as a result of quantitative assessment of STIR (short tau inversion recovery)
routine post-­treatment CT studies.34 The role of cross-­sectional MRI to investigate nodal involvement in 35 patients, specifically
imaging in the surveillance of patients post-­oesophagectomy looking at the signal intensity ratio between each lymph node
is contentious and further research is required to standardise and the normal oesophageal wall. Compared to histology, 152
practice. of 482 nodes were detected by MRI, with the signal intensity
18 ratio significantly higher in pathological nodes.44 MRI enhanced
F-­FDG PET/CT has the highest sensitivity for recurrent disease with superparamagnetic iron oxide nanoparticle (SPIO) contrast
at 89–100%, although specificity is much lower (55–94%).33 Local has the potential to differentiate normal or reactive from meta-
inflammation in the oesophagus can cause false positive FDG-­ static lymph nodes because normal nodes contain substantial
uptake and should be confirmed with endoscopy.35 The radia- phagocytosed SPIO, which appears low signal due to magnetic
tion doses associated with PET/CT mean it is currently used as susceptibility and T2* shortening, whereas malignant nodes
a problem-­solving tool in the setting of indeterminate findings are intermediate signal.45 In oesophageal cancer, one study of
on CT, rather than as part of a routine surveillance programme. nine patients showed this to be a feasible technique,46 although
concerns around safety in clinical practice have been raised.47
STATE-OF-THE-ART IMAGING IN STAGING
The combined limitations of CT, 18F-­FDG PET/CT and EUS
mean that new imaging technologies are needed to improve the Whole-body MRI
delineation of disease extent, the detection of lymph node metas- Following the success of the STREAMLINE-­C trial in colorectal
tases, and the assessment of treatment response. There is an cancer,48 the role of whole body-­ MRI in oesophago-­ gastric
opportunity for state-­of-­the-­art imaging techniques to address cancer staging has been raised. Whole-­body MRI might allow
these gaps, ensuring that patients are stratified to the most appro- a more streamlined and cost-­effective staging pathway, with
priate treatment. local and distant staging of disease having equivalent accuracy
to current investigations in a single radiological examination.
MRI However, high-­quality studies in oesophago-­gastric cancer are
MRI has excellent soft tissue contrast and can identify the lacking. In one small study, whole-­body MRI had a reported
normal layers of the oesophageal wall (Figure  3), with poten- equivalent accuracy to PET/CT for N-­stage and excluding meta-
tial to improve stratification of patients towards endoscopic static disease (n = 49 patients, using surgical specimens or EUS
resection, upfront surgery or neoadjuvant therapy. In seminal for N-­stage, and metastases being detected in only two patients
work, Riddell et al showed T2 weighted images from 1.5 T MRI and with both modalities).43 This has not yet been tested within
were comparable to EUS in differentiating T2 from T3, but a rigorous clinical trial. In systematic review, albeit in gastric
overstaged T1 tumours. Overall, 83% (n = 28/37) were staged cancer, whole-­body MRI was comparable to CT in detection
correctly against histology; 16% were overstaged and 8% under- of peritoneal disease, which is a common metastatic site from
staged.36 Ex-­vivo studies at ultra-­
high field strength (4.7 T oesophageal cancer.49,50 The addition of DWI in a whole-­body
and 7 T) have shown up to 100% accuracy for T- and N-­stage, MRI protocol improves sensitivity over morphological images
although application to clinical practice has not been tested.37 alone (90% vs  73% in assessment of 255 peritoneal deposits in 34

4 of 12 birpublications.org/bjr Br J Radiol;95:20220410
 State-­of-­the-­art imaging in oesophago-­gastric cancer BJR

Figure 4. Axial fused 18F-­FDG PET/MRI image showing FDG preferences, age and pathology. Neoadjuvant therapy prior to
uptake in the primary tumour, locoregional nodes and liver resection is the standard of care for patients with Stage II to III
metastasis in a patient with Stage IV oesophago-­gastric can- disease, offering a 5.1% absolute survival advantage at 2 years
cer (Images courtesy of Professors G Cook and V Goh, St. vs surgery alone in adenocarcinomas, highlighting the impor-
Thomas’ Hospital). FDG, fluorodeoxyglucose. tance of accurate staging.56 However, only a minority of patients
(14.8%) demonstrate a good response to treatment, defined as
a tumour regression grade (TRG) of 1 or 2.57 Further, disease
length is an important consideration when planning opera-
tion type or suitability for radiotherapy, the latter being depen-
dent on a maximum field length of approximately 12 cm.58 CT
is traditionally used to define the tumour and organs at risk
during radiotherapy planning, however, 18F-­FDG PET/CT has
now been incorporated into the planning process and uses the
metabolic activity of the primary tumour and nodes to adjust the
irradiated volumes.59 When performed, EUS provides measure-
ments defining the location of disease and important anatomical
landmarks such as the aortic arch, carina and diaphragm which
can be used to augment the radiotherapy plan (Figure 5). There
is an opportunity for state-­of-­the-­art imaging and biomarkers to
improve patient selection for individualised treatment, enabling
prediction of the likelihood of response, expediting patients for
surgery who are unlikely to respond, and improved definition of
disease.

STATE-OF-THE-ART IMAGING IN PREDICTING

RESPONSE AND OUTCOME
patients, verified at laparoscopy and with histology, but again not
in oesophageal cancer).51 Diffusion-weighted MRI
Apparent diffusion coefficient (ADC) values are lower in areas of
Integrated PET/MRI increased cellularity such as tumour. There have been conflicting
18
F-­FDG PET/MRI combines the benefits of both 18F-­FDG PET/ reports regarding the predictive value of pre-­treatment ADC
CT and MRI, i.e. optimising locoregional and distant staging in values. Aoyagi et al. (n = 80; SCC undergoing chemoradio-
one examination (Figure 4). Lee et al found a diagnostic accuracy therapy) reported that higher pre-­treatment ADC values (1.1 ×
of 67% for T-­stage compared to 87% for EUS in a prospective 10−3 mm2/s and greater) were associated with increased likeli-
study of 19 patients with pathological correlation. 18F-­FDG PET/ hood of response assessed by RECIST (hazard ratio 23.4 (6.1–
MRI N-­stage accuracy was 87% compared to 67% for EUS, and 89.6)) and increased survival (42% 1 year survival versus 18%)
50% for CT.52 Good agreement between 18F-­FDG PET/MRI and in advanced SCC,60 whereas De Cobelli et al. (n = 32 oesoph-
PET/CT for N- and M-­stage has been reported.53 Sharkey et al ageal and gastric cancers; 81% adenocarcinoma; undergoing
compared TNM stage between tumour board consensus (from either chemoradiotherapy or chemotherapy, respectively) found
all available diagnostic tests, excluding PET/MRI), 18F-­FDG responders to neoadjuvant therapy assessed by Mandard TRG
PET/CT alone, and 18F-­FDG PET/MRI alone. In this prospec- had significantly lower pre-­treatment ADC values (mean ADC
tive study of 22 patients, 10 with metastatic disease, additional in responders (TRG 1–3) was 1.32 ± 0.33 x 10−3mm2/s vs 1.63 ±
metastases were found on 18F-­FDG PET/MRI in 30% of cases 0.41 x 10−3mm2/s in non-­responders. p = 0.002).61
compared to PET/CT (two peritoneal and one liver), which has
potential clinical relevance when detection of a single metastatic Perfusion imaging
site can change management between curative and palliative Perfusion parameters from DCE-­ MRI or perfusion CT are
therapy.54 18F-­FDG PET/MRI can be time-­saving compared to a further potential avenue to predict response. The transfer
acquiring the images separately,55 however, this strategy assumes constant (Ktrans) is the rate of leakage of contrast into the inter-
that patients will require a PET examination during their staging. stitium and is related to blood flow and tissue permeability
Potential imaging biomarkers from 18F-­FDG PET/CT (glucose (Figure  6). In many studies, Ktrans is the DCE-­MRI parameter
metabolism) and MRI (perfusion phenotype from dynamic which is most predictive of response to treatment. Higher
contrast-­enhanced (DCE) MRI or cellularity from DWI) are pre-­treatment Ktrans values have been associated with better
discussed below. Acquisition of these modalities contemporane- response to neoadjuvant and palliative chemoradiotherapy (Sun
ously could be advantageous. et al n = 59 SCC and Lei et al n = 25 SCC).62,63 Heethuis et al.
(n = 25, 84% adenocarcinoma) found significantly different
STATE-OF-THE-ART IMAGING IN TREATMENT pre-­treatment MRI perfusion parameters (25% percentile of
PLANNING iAUC) in good pathological responders.64 Using perfusion CT,
In oesophago-­gastric cancer, treatment decisions are based on several authors have found increased blood flow associated with
a number of factors including disease stage, patient fitness and an increased likelihood of response. Hayano et al (n = 31 SCC)

5 of 12 birpublications.org/bjr Br J Radiol;95:20220410
 BJR Withey et al

Figure 5. CT-­based radiotherapy planning image of a distal oesophageal tumour demonstrating the GTV (red), CTV (pink) and
PTV (blue) with several isodose lines contoured by clinical oncologists during target volume delineation (Images courtesy of Dr
Owen Nicholas, South West Wales Cancer Centre). CTV, clinical target volume; GTV, gross tumour volume; PTV, planning target
volume.

reported mean blood flow 90.2 ml/100 g/min in responders, PET/CT

defined by >50% tumour volume decrease, vs 35.9 ml/100 g/ Published data on the value of 18F-­FDG PET/CT parameters
min in non-­responders (p = 0.0004).65 Zhao et al. (n = 27 SCC) in prediction of response or survival are heterogeneous and
found responders, assessed by RECIST, had mean blood flow conflicting. Authors have reported that increased maximum
42.1 vs 27.5 ml/100 g/min (p = 0.007) for non-­responders.65,66 It standardised uptake value (SUVmax) is associated with increased
is hypothesised that better perfused tumours may receive higher likelihood of response67 whilst others found the opposite68 or no
doses of chemotherapeutic agents which has direct relevance for relationship at all.69
treatment selection.

Figure 6. Axial fat-­saturated T1 weighted MRI performed before (a), 30 s (b), 70 s (c) and 120 s (d) after contrast administration in
a patient with a distal oesophageal tumour (arrows). Contrast-­enhancement curve showing a type-­3 washout curve (e). Focussed
Ktrans map showing mean Ktrans value of 0.39 min−1 within the oesophageal tumour (arrow) (f). (Images courtesy of Professor V
Goh, St. Thomas’ Hospital).

6 of 12 birpublications.org/bjr Br J Radiol;95:20220410
 State-­of-­the-­art imaging in oesophago-­gastric cancer BJR

In addition to 18F-­FDG, there are opportunities for use of novel Diffusion-weighted MRI
PET tracers in oesophago-­ gastric cancer. Human epidermal MRI does not involve ionising radiation, therefore can be used
growth receptor 2 (HER2-­ neu) is overexpressed in many repeatedly throughout the staging and treatment pathway.
oesophageal adenocarcinomas. Monoclonal antibodies targeting Changes in tumour ADC values during neoadjuvant treat-
this receptor have been developed as immunotherapy-­ based ment have been reported as a potential method of predicting
treatments, e.g. Trastuzumab, which has been shown to improve response with early increases in ADC associated with favour-
overall survival in the metastatic setting.70 However, heteroge- able response (Figure  7).75,76 van Rossum et al (n = 20; 75%
neity of metastatic sites is not a unique problem to oesophago-­
adenocarcinoma) reported that less than 29% ADC increase had
gastric cancer and can lead to treatment failure. In a study of 33
100% sensitivity for residual tumour at the end of neoadjuvant
patients undergoing both 89Zr-­Trastuzumab PET/CT and 18F-­
therapy.76 Heethuis et al (n = 45; 84% adenocarcinoma) reported
FDG PET/CT, an average of 5.5 lesions per patients were iden-
tified using 89Zr-­Trastuzumab PET/CT compared to 8 using that ADC values increased during treatment in both good and
18
F-­FDG PET/CT.71 This shows the potential of novel imaging poor responder groups, although the change was more marked
tracers to promote and select patients for individualised treat- in good responders (23.5 ± 20.5% and 9.8 ± 11.7%, respec-
ment plans. Further, the possibility of theranostic treatment tively, p = 0.035).77 Borggreve et al performed DWI before and
options using 177Lu-­ Trastuzumab are being researched for weekly during neoadjuvant chemoradiotherapy to determine the
HER2 positive breast cancer,72 which may translate to oesophago-­ optimum time to predict complete pathological response (TRG
gastric cancer. Potential other molecular targets include vascular 1). Seven of 24 patients had complete response. The percentage
endothelial growth factor, epidermal growth factor receptor73 increase in ADC from baseline to the 2-­week scan was most
and hypoxia imaging.74 predictive of subsequent complete response (mean increase 36%
in complete response vs 16% in non-­complete response), with a
State-of-the-art imaging in assessing response c-­index of 0.87, which increased to 0.97 after exclusion of small
Conventional imaging cannot adequately predict treatment tumours.78 Visual assessment on post-­treatment DWI had good
response with sufficient accuracy to change current clinical path- sensitivity for detection of residual local disease, but low speci-
ways. There are opportunities for state-­of-­the-­art imaging tech- ficity (42–50%).79 Some authors have found an inverse relation-
niques to meet this challenge and improve treatment response ship between ADC values and TRG, with higher post-­treatment
assessment. A further role of imaging could be to predict ADC values associated with better response (de Cobelli et al n =
response at an earlier time point during neoadjuvant therapy, 32 oesophageal and gastric cancers; and more recently Giganti et
to allow changes to the treatment regimen, or an alternative al n = 18 oesophageal, including some of the same patients as the
approach. study by de Cobelli et al).38,61

Figure 7. Axial T2 weighted MRI, B900 diffusion-­weighted MRI, and apparent diffusion coefficient map before (a, b, c) and after
(d, e, f) neoadjuvant chemoradiotherapy. The tumour is at the gastro-­oesophageal junction, above a moderate-­sized hiatus her-
nia. The tumour decreased in bulk, and mean tumour apparent diffusion coefficient increased from 800 × 10−3 mm2/s to 1850 ×
10−3 mm2/s (Images courtesy of Professor V Goh, St. Thomas’ Hospital).

7 of 12 birpublications.org/bjr Br J Radiol;95:20220410
 BJR Withey et al

Dual energy CT study used early PET/CT during neoadjuvant chemotherapy to

Dual energy CT can provide quantitative information about identify metabolic non-­responders. Patients with greater than 35%
iodine contrast uptake. A full review of different dual energy CT decrease in SUVmax continued with the standard pathway, but those
methodologies is described by McCollough et al80 but in brief it with less than 35% decrease were considered non-­responders and
utilises varying X-­ray energy spectra to simultaneously (either proceeded straight to surgery, showing the potential of PET/CT to
with two separate X-­ray sources or with rapid kV switching) or alter the neoadjuvant treatment pathway. Of 54 patients classified
sequentially acquire two sets of image data at the same anatom- as non-­responders on early PET/CT, none had a major histological
ical location. This can provide more information on the constit- response (<10% viable tumour remaining) at the time of resection.87
uent materials present within a voxel as the images are combined Metabolic responders had better progression-­free survival (29.7 vs
with spectral characterisation.81 In a single centre study of 45 14.1 months). To determine the effect of early PET/CT for treat-
patients, iodine concentrations after chemoradiotherapy were ment change, it would be necessary to randomise metabolic non-­
significantly lower in patients that responded vs non-­responders responders to either continued standard care or early surgery, which
(assessed by RECIST v 1.1).82 is currently lacking and may not be feasible.

18
Perfusion imaging F-­FDG PET/CT performed after neoadjuvant therapy can be chal-
Decreased blood flow65,83 and increased mean transit time84 have lenging to interpret as radiation-­induced oesophagitis or ulceration
been associated with improved response to chemoradiotherapy is common and also has FDG uptake, although this will generally
and increased survival. In a single centre study of 40 SCC patients resolve within 6–12 weeks of radiotherapy.88 Despite PET/CT having
following neoadjuvant therapy, Djuric-­ Stefanovic et al found the highest sensitivity for identifying distant metastases, up to 5%
increased blood flow and increased blood volume were associated of patients will have false-­positive sites of uptake requiring further
with higher Mandard TRG scores (i.e. worse response). They found imaging or biopsy.89 Equally, a lack of FDG uptake at the primary
blood flow <30 ml/min/100g corresponded with complete patholog- site cannot distinguish between microscopic residual disease and
ical response.85 complete pathological response.90 In one study of 135 patients
with negative biopsies after neoadjuvant treatment, 85 had residual
Using DCE-­MRI, Heethuis et al (n = 45, 84% adenocarcinoma) tumour at oesophagectomy.91 As part of the prospective preSANO
found the change in iAUC (the initial area under the Gadolinium trial, 129 patients underwent 18F-­FDG PET/CT after neoadjuvant
concentration curve, reflecting contrast inflow and vascular leakage) chemoradiotherapy and prior to surgery. They were able to identify
between pre-­treatment MRI and MRI performed after one cycle of complete response (TRG 1) with 80% sensitivity and 37% specificity.
neoadjuvant therapy could discriminate between TRG 1–2 and 3–5 15% of patients with a poor response (TRG 3 or 4) had a complete
(10.6±17.6% increase in iAUC in complete responders vs 45.2 ± metabolic response on PET/CT.23 A study by Cerfolio et al used an
41.5% increase in non-­complete responders. p = 0.028).77 In oesoph- SUVmax cut-­off of 3.25 and found this predicted complete response
ageal SCC (n = 59) treated with chemoradiotherapy and assessed with with 67% sensitivity and specificity.92 Further, FDG PET/CT was
DCE-­MRI before and after treatment, the absolute post-­treatment significantly better at predicting complete response than EUS and CT
Ktrans (the transfer constant, reflecting the rate of leakage of Gado- (89% vs 67 and 71%, respectively).92
linium from vessels to the interstitium) values were lower in complete
responders (by RECIST v 1.1) vs the remainder, and the decreases Novel PET tracers could play a potential role in predicting response
in Ktrans and Kep (the rate constant, reflecting the rate of return of by investigating changes in uptake values early during treatment.
18
contrast back from the interstitium) values were significantly greater F-­Fluorothymidine is a PET tracer that acts as a marker of cellular
in complete responders.63 Other studies (n = 25 SCC) found signifi- proliferation and is thought to discriminate between tumour and
cantly lower Ktrans and higher Ve (the relative volume of extracellular inflammation. Chen et al (n = 34 SCC) found a decrease in 18F-­Fluo-
extravascular space available to accumulate Gadolinium) in complete rothymidine uptake 4 weeks after starting neoadjuvant chemoradio-
responders vs partial response when assessed by RECIST v 1.1.62 therapy for SCC was associated with improved locoregional control
Conversely, Heethuis et al did not find post-­treatment iAUC was able and better 2 year progression-­free survival, whereas 18F-­FDG uptake
to differentiate complete pathological responders vs non-­complete was not predictive.93 In a pilot study of 26 patients undergoing both
18
responders.77 A combination of DCE-­MRI (change in Ktrans between F-­FDG PET/CT and 11C-­thiothymidine before and after neoad-
per- and pre-­treatment) and DW-­MRI (change in ADC between juvant treatment, lower post-­treatment 11C-­thiothymidine SUVmax
post- and pre-­treatment) has yielded better performance than either and a greater percentage decrease in 11C-­thiothymidine SUVmax
individually (c-­index 0.89 vs 0.79 and 0.75, respectively).77 were associated with pathological response. Similarly, the percentage
change in 18F-­FDG SUVmax was also associated with pathological
response, however absolute 18F-­FDG SUVmax was not.94
PET/CT
Published data on the value of 18F-­ FDG PET/CT parameters
for predicting response or survival are heterogeneous. PET/CT State-of-the-art imaging of recurrent disease
performed during treatment has been shown to predict response as Only small studies have compared MRI vs CT in the setting of
metabolic changes can precede pathological response. In a 2017 meta-­ recurrent disease. In 23 patients with recurrent disease following
analysis including 13 studies and 697 patients, 8 studies found some oesophagectomy, CT and MRI performed equally well at iden-
predictive value in performing PET/CT during neo-­adjuvant chemo- tifying intraluminal local recurrence, liver metastases, and
radiotherapy, whereas 5 did not. Pooled sensitivity for complete malignant pleural and pericardial effusions. MRI identified
pathological response was 63–100% and specificity 50–76%.86 One more bone metastases and was superior in identifying malignant

8 of 12 birpublications.org/bjr Br J Radiol;95:20220410
 State-­of-­the-­art imaging in oesophago-­gastric cancer BJR

oesophageal wall thickening but performed worse in assessment ACKNOWLEDGEMENTS

of lung metastases.95 Shuto et al investigated DWI in assessment SJW was supported by a Cancer Research UK Research Bursary
of nodal relapse (n = 47 suspected nodes identified on CT, with (C66940/A28141). VG acknowledges funding support from
histology as reference standard), finding restricted diffusion Cancer Research UK OCCAMS (Oesophageal Cancer Clinical
assessed on ADC map had diagnostic accuracy of 81% compared and Molecular Stratification Consortium) programme grant
to 87% from 18F-­FDG PET/CT.96 (RG66287); Cancer Research UK National Imaging Transla-
tional Accelerator Award (C4278/A27066); Wellcome EPSRC
Centre for Medical Engineering at King’s College London (WT
CONCLUSIONS 203148/Z/16/Z); the UKRI Medical Imaging and Artificial Intel-
State-­of-­the-­art imaging techniques have the potential to trans- ligence Centre for Value Based Healthcare; and Department
form the diagnostic, staging and treatment pathway for patients of Health via the NIHR Comprehensive Biomedical Research
with oesophago-­gastric cancer. Whole body MRI, PET/MRI, and Centre award to Guy’s & St Thomas' NHS Foundation Trust in
novel PET tracers have shown promise in early research studies partnership with King’s College London/King’s College Hospital
of oesophago-­gastric cancer and may allow more precise delinea- NHS Foundation Trust. KGF is supported by research funding
tion of disease extent and prediction of treatment response thus from Health and Care Research Wales (HCRW) and the Moon-
optimising treatment decisions and patient outcomes. dance Foundation at Velindre Cancer Centre.

REFERENCES
1. Cancer Research. Oesophageal cancer 8. Boys JA, Worrell SG, Chandrasoma P, 14. Berger AC, Scott WJ. Noninvasive staging of
incidence statistics. Available from: https:// Vallone JG, Maru DM, Zhang L, et al. Can esophageal carcinoma. J Surg Res 2004; 117:
www.cancerresearchuk.org/health-​ the risk of lymph node metastases be gauged 127–33. https://doi.org/10.1016/j.jss.2003.12.​
professional/cancer-statistics/statistics-by-​ in endoscopically resected submucosal 023
cancer-type/oesophageal-cancer/incidence esophageal adenocarcinomas? A multi-­ 15. Guo J, Wang Z, Qin J, Zhang H, Liu W,
2. Sung H, Ferlay J, Siegel RL, Laversanne M, center study. J Gastrointest Surg 2016; 20: Zhao Y, et al. A prospective analysis of the
Soerjomataram I, Jemal A, et al. Global 6–12; . https://doi.org/10.1007/s11605-015-​ diagnostic accuracy of 3 T MRI, CT and
cancer statistics 2020: GLOBOCAN 2950-9 endoscopic ultrasound for preoperative T
estimates of incidence and mortality 9. van Hagen P, Hulshof MCCM, van Lanschot staging of potentially resectable esophageal
worldwide for 36 cancers in 185 countries. JJB, Steyerberg EW, van Berge Henegouwen cancer. Cancer Imaging 2020; 20: 64. https://​
CA Cancer J Clin 2021; 71: 209–49. https://​ MI, Wijnhoven BPL, et al. Preoperative doi.org/10.1186/s40644-020-00343-w
doi.org/10.3322/caac.21660 chemoradiotherapy for esophageal or 16. van Vliet EPM, Heijenbrok-­Kal MH, Hunink
3. Rubenstein JH, Shaheen NJ. Epidemiology, junctional cancer. N Engl J Med 2012; MGM, Kuipers EJ, Siersema PD. Staging
diagnosis, and management of esophageal 366: 2074–84. https://doi.org/10.1056/​ investigations for oesophageal cancer: a
adenocarcinoma. Gastroenterology 2015; 149:
NEJMoa1112088 meta-­analysis. Br J Cancer 2008; 98: 547–57.
10. Deng J, Chu X, Ren Z, Wang B. Relationship https://doi.org/10.1038/sj.bjc.6604200
S0016-­5085(15)00642-­3: 302-17. . https://doi.​
between T stage and survival in distantly 17. Thosani N, Singh H, Kapadia A, Ochi N, Lee
org/10.1053/j.gastro.2015.04.053
metastatic esophageal cancer: A STROBE-­ JH, Ajani J, et al. Diagnostic accuracy of EUS
4. Kauppila JH, Mattsson F, Brusselaers N,
compliant study. Medicine (Baltimore) 2020; in differentiating mucosal versus submucosal
Lagergren J. Prognosis of oesophageal
99: e20064: 19: . https://doi.org/10.1097/MD.​ invasion of superficial esophageal cancers:
adenocarcinoma and squamous cell
0000000000020064 a systematic review and meta-­analysis.
carcinoma following surgery and no surgery
11. Chiapponi C, Berlth F, Plum PS, Betzler C, Gastrointest Endosc 2012; 75: 242–53. https://​
in a nationwide swedish cohort study. BMJ
Stippel DL, Popp F, et al. Oligometastatic doi.org/10.1016/j.gie.2011.09.016
Open 2018; 8(5): e021495. https://doi.org/10.​
disease in upper gastrointestinal cancer - 18. Vazquez-­Sequeiros E, Wiersema MJ, Clain
1136/bmjopen-2018-021495
how to proceed? Visc Med 2017; 33: 31–34. JE, Norton ID, Levy MJ, Romero Y, et al.
5. Daly JM, Fry WA, Little AG, Winchester DP,
https://doi.org/10.1159/000455683 Impact of lymph node staging on therapy
McKee RF, Stewart AK, et al. Esophageal 12 . National Institute for Health and Care of esophageal carcinoma. Gastroenterology
cancer: results of an american college of Excellence. Oesophago-­gastric cancer: 2003; 125: 1626–35. https://doi.org/10.1053/​
surgeons patient care evaluation study. J Am assessment and management in adults j.gastro.2003.08.036
Coll Surg 2000; 190: 562–72; . https://doi.org/​ (NG83). https://www.nice.org.uk/guidance/​ 19. Morgan MA, Twine CP, Lewis WG, Lambe R,
10.1016/s1072-7515(00)00238-6 ng83/resources/oesophagogastric-cancer-​ Oliphant HE, Robinson M, et al. Prognostic
6. American Joint Committee on Cancer. AJCC assessment-and-management-in-adults-pdf-​ significance of failure to cross esophageal
Cancer Staging Manual. 8th Edition ed. 18376930144692018. tumors by endoluminal ultrasound. Dis
Springer International Publishing; 2017. 13. Allum WH, Blazeby JM, Griffin SM, Esophagus 2008; 21: 508–13. https://doi.org/​
7. Glickman JN. Section II: pathology and Cunningham D, Jankowski JA, Wong R, 10.1111/j.1442-2050.2008.00809.x
pathologic staging of esophageal cancer. et al. Guidelines for the management of 20. Patel N, Foley KG, Powell AG, Wheat JR,
Semin Thorac Cardiovasc Surg 2003; 15: oesophageal and gastric cancer. Gut 2011; 60: Chan D, Fielding P, et al. Propensity score
167–79. https://doi.org/10.1016/S1043-0679(​ 1449–72. https://doi.org/10.1136/gut.2010.​ analysis of 18-­FDG PET/CT-­enhanced
03)70025-2 228254 staging in patients undergoing surgery for

9 of 12 birpublications.org/bjr Br J Radiol;95:20220410
 BJR Withey et al

esophageal cancer. Eur J Nucl Med Mol therapy for oesophageal cancer. Br J Surg. 39. Lee SL, Yadav P, Starekova J, Christensen L,
Imaging 2019; 46: 801–9. https://doi.org/10.​ 2017;104(4):408-17. Chandereng T, Chappell R, et al. Diagnostic
1007/s00259-018-4118-9 30. Foley KG, Christian A, Fielding P, Lewis performance of MRI for esophageal
21. You JJ, Wong RKS, Darling G, Gulenchyn WG, Roberts SA. Accuracy of contemporary carcinoma: A systematic review and meta-­
K, Urbain J-­L, Evans WK. Clinical utility of oesophageal cancer lymph node staging with analysis. Radiology 2021; 299: 583–94.
18F-­fluorodeoxyglucose positron emission radiological-­pathological correlation. Clin https://doi.org/10.1148/radiol.2021202857
tomography/computed tomography in the Radiol 2017; 72: e1-7. https://doi.org/10.​ 40. Qu J, Zhang H, Wang Z, Zhang F, Liu H,
staging of patients with potentially resectable 1016/j.crad.2017.02.022 Ding Z, et al. Comparison between free-­
esophageal cancer. J Thorac Oncol 2013; 31. Mariette C, Balon JM, Piessen G, Fabre S, breathing radial VIBE on 3-­T MRI and
8: 1563–69. https://doi.org/10.1097/JTO.​ Van Seuningen I, Triboulet JP. Pattern of endoscopic ultrasound for preoperative T
0000000000000006 recurrence following complete resection of staging of resectable oesophageal cancer,
22. Kumar P, Damle NA, Bal C. Role of F18-­ esophageal carcinoma and factors predictive with histopathological correlation. Eur
FDG PET/CT in the staging and restaging of recurrent disease. Cancer 2003; 97: Radiol 2017; 28: 780–87. https://doi.org/10.​
of esophageal cancer: A comparison with 1616–23. https://doi.org/10.1002/cncr.11228 1007/s00330-017-4963-0
CECT. Indian J Surg Oncol 2011; 2: 343–50. 32. Blom RL, Lagarde SM, van Oudenaarde 41. van Rossum PSN, van Lier ALHMW, Lips
https://doi.org/10.1007/s13193-012-0128-4 K, Klinkenbijl JHG, Hulshof MC, van IM, Meijer GJ, Reerink O, van Vulpen M,
23. Valkema MJ, Noordman BJ, Wijnhoven Laarhoven HW, et al. Survival after recurrent et al. Imaging of oesophageal cancer with
BPL, Spaander MCW, Biermann K, Lagarde esophageal carcinoma has not improved over FDG-­PET/CT and MRI. Clin Radiol 2015;
SM, et al. Accuracy of 18F-­FDG PET/CT in the past 18 years. Ann Surg Oncol 2013; 20: 70: S0009-­9260(14)00379-­1: 81–95: . https://​
predicting residual disease after neoadjuvant 2693–98. https://doi.org/10.1245/s10434-​ doi.org/10.1016/j.crad.2014.07.017
chemoradiotherapy for esophageal cancer. J 013-2936-3 42. Dave UR, Williams AD, Wilson JA, Amin Z,
Nucl Med 2019; 60: 1553–59. https://doi.org/​ 33. Goense L, van Rossum PSN, Reitsma JB, Gilderdale DJ, Larkman DJ, et al. Esophageal
10.2967/jnumed.118.224196 Lam MGEH, Meijer GJ, van Vulpen M, et al. cancer staging with endoscopic MR imaging:
24. Meredith KL, Weber JM, Turaga KK, Siegel Diagnostic performance of 18Diagnostic pilot study. Radiology 2004; 230: 281–86.
EM, McLoughlin J, Hoffe S, et al. Pathologic performance of. J Nucl Med 2015; 56: https://doi.org/10.1148/radiol.2301021047
response after neoadjuvant therapy is the 995–1002. https://doi.org/10.2967/jnumed.​ 43. Malik V, Harmon M, Johnston C, Fagan
major determinant of survival in patients 115.155580 AJ, Claxton Z, Ravi N, et al. Whole body
with esophageal cancer. Ann Surg Oncol 34. Lou F, Sima CS, Adusumilli PS, Bains MS, MRI in the staging of esophageal cancer-
2010; 17: 1159–67. https://doi.org/10.1245/​ Sarkaria IS, Rusch VW, et al. Esophageal -A prospective comparison with whole
s10434-009-0862-1 cancer recurrence patterns and implications body 18F-­FDG PET-­CT. Dig Surg 2015; 32:
25. Berger AC, Farma J, Scott WJ, Freedman for surveillance. J Thorac Oncol 2013; 8: 397–408. https://doi.org/10.1159/000431292
G, Weiner L, Cheng JD, et al. Complete 1558–62. https://doi.org/10.1097/01.JTO.​ 44. Alper F, Turkyilmaz A, Kurtcan S, Aydin Y,
response to neoadjuvant chemoradiotherapy 0000437420.38972.fb Onbas O, Acemoglu H, et al. Effectiveness
in esophageal carcinoma is associated with 35. Park SY, Lee HS, Jang HJ, Lee JY, Joo J, Zo JI. of the STIR turbo spin-­echo sequence MR
significantly improved survival. J Clin Oncol The role of one-­year endoscopic follow-­up imaging in evaluation of lymphadenopathy
2005; 23: 4330–37. https://doi.org/10.1200/​ for the esophageal remnant and gastric in esophageal cancer. Eur J Radiol 2011; 80:
JCO.2005.05.017 conduit after esophagectomy with gastric 625–28. https://doi.org/10.1016/j.ejrad.2010.​
26 . Brücher BL, Becker K, Lordick F, Fink U, reconstruction for esophageal squamous cell 08.003
Sarbia M, Stein H, et al. The clinical impact carcinoma. Yonsei Med J 2013; 54: 381–88. 45. Pannu HK, Wang KP, Borman TL, Bluemke
of histopathologic response assessment https://doi.org/10.3349/ymj.2013.54.2.381 DA. MR imaging of mediastinal lymph
by residual tumor cell quantification in 36 . Riddell AM, Allum WH, Thompson JN, nodes: evaluation using a superparamagnetic
esophageal squamous cell carcinomas. Wotherspoon AC, Richardson C, Brown G. contrast agent. J Magn Reson Imaging 2000;
Cancer. 2006;106(10):2119-27. The appearances of oesophageal carcinoma 12: 899–904. https://doi.org/10.1002/1522-​
27. Noordman BJ, Spaander MCW, Valkema R, demonstrated on high-­resolution, T2-­ 2586(200012)12:6<899::aid-jmri13>3.0.co;​
Wijnhoven BPL, Berge Henegouwen MI, weighted MRI, with histopathological 2-r
Shapiro J, et al. Detection of residual disease correlation. European radiology. 46. Pultrum BB, van der Jagt EJ, van Westreenen
after neoadjuvant chemoradiotherapy for 2007;17(2):391-9. HL, van Dullemen HM, Kappert P, Groen
oesophageal cancer (presano): a prospective 37 . Yamada I, Hikishima K, Miyasaka N, H, et al. Detection of lymph node metastases
multicentre, diagnostic cohort study. Lancet Kawano T, Tokairin Y, Ito E, et al. Esophageal with ultrasmall superparamagnetic iron
Oncol 2018; 19: 965–74. https://doi.org/10.​ carcinoma: ex vivo evaluation with diffusion-­ oxide (USPIO)-­enhanced magnetic
1016/S1470-2045(18)30201-8 tensor MR imaging and tractography at 7 T. resonance imaging in oesophageal cancer:
28 . Beseth BD, Bedford R, Isacoff WH, Radiology. 2014;272(1):164-73. a feasibility study. Cancer Imaging 2009; 9:
Holmes EC, Cameron RB. Endoscopic 38. Giganti F, Ambrosi A, Petrone MC, Canevari 19–28. https://doi.org/10.1102/1470-7330.​
ultrasound does not accurately assess C, Chiari D, Salerno A, et al. Prospective 2009.0004
pathologic stage of esophageal cancer after comparison of MR with diffusion-­weighted 47. Singh N, Jenkins GJS, Asadi R, Doak SH.
neoadjuvant chemoradiotherapy. Am Surg. imaging, endoscopic ultrasound, MDCT Potential toxicity of superparamagnetic iron
2000;66(9):827-31. and positron emission tomography-­CT in oxide nanoparticles (SPION). Nano Rev
29 . Findlay JM, Bradley KM, Wang LM, the pre-­operative staging of oesophageal 2010; 1. https://doi.org/10.3402/nano.v1i0.​
Franklin JM, Teoh EJ, Gleeson FV, cancer: results from a pilot study. Br J Radiol 5358
et al. Metabolic nodal response as a 2016; 89: 1068. https://doi.org/10.1259/bjr.​ 48. Taylor SA, Mallett S, Beare S, Bhatnagar G,
prognostic marker after neoadjuvant 20160087 Blunt D, Boavida P, et al. Diagnostic accuracy

10 of 12 birpublications.org/bjr Br J Radiol;95:20220410
 State-­of-­the-­art imaging in oesophago-­gastric cancer BJR

of whole-­body MRI versus standard imaging validate pathological assessment of response 65. Hayano K, Okazumi S, Shuto K, Matsubara
pathways for metastatic disease in newly to neoadjuvant therapy in oesophagogastric H, Shimada H, Nabeya Y, et al. Perfusion CT
diagnosed colorectal cancer: the prospective adenocarcinoma. Br J Surg 2017; 104: can predict the response to chemoradiation
streamline C trial. Lancet Gastroenterol 1816–28. https://doi.org/10.1002/bjs.10627 therapy and survival in esophageal squamous
Hepatol 2019; 4: S2468-­1253(19)30056-­1: 58. Crosby T, Hurt CN, Falk S, Gollins cell carcinoma: initial clinical results. Oncol
529–37: . https://doi.org/10.1016/S2468-​ S, Mukherjee S, Staffurth J, et al. Rep 2007; 18: 901–8. https://doi.org/10.3892/​
1253(19)30056-1 Chemoradiotherapy with or without or.18.4.901
49. Laghi A, Bellini D, Rengo M, Accarpio cetuximab in patients with oesophageal 66. Zhao K, Wang C, Mao Q, Shang D, Huang Y,
F, Caruso D, Biacchi D, et al. Diagnostic cancer (SCOPE1): a multicentre, phase 2/3 Ma L, et al. The flow-­metabolism ratio might
performance of computed tomography and randomised trial. The Lancet Oncology 2013; predict treatment response and survival in
magnetic resonance imaging for detecting 14: 627–37. https://doi.org/10.1016/S1470-​ patients with locally advanced esophageal
peritoneal metastases: systematic review and 2045(13)70136-0 squamous cell carcinoma. EJNMMI Res 2020;
meta-­analysis. Radiol Med 2017; 122: 1–15. 59. Ng SP, Tan J, Osbourne G, Williams L, 10(1): 57. https://doi.org/10.1186/s13550-​
https://doi.org/10.1007/s11547-016-0682-x Bressel MAB, Hicks RJ, et al. Follow up 020-00647-9
50. Wang Z, Chen JQ. Imaging in assessing results of a prospective study to evaluate 67. Javeri H, Xiao L, Rohren E, Lee JH, Liao Z,
hepatic and peritoneal metastases of the impact of FDG-­PET on CT-­based Hofstetter W, et al. The higher the decrease
gastric cancer: a systematic review. BMC radiotherapy treatment planning in the standardized uptake value of positron
Gastroenterol 2011; 11: 19. https://doi.org/10.​ for oesophageal cancer. Clinical and emission tomography after chemoradiation,
1186/1471-230X-11-19 Translational Radiation Oncology 2017; 2: the better the survival of patients with
51. Low RN, Sebrechts CP, Barone RM, Muller 76–82. https://doi.org/10.1016/j.ctro.2017.01.​ gastroesophageal adenocarcinoma. Cancer
W. Diffusion-­weighted MRI of peritoneal 008 2009; 115: 5184–92. https://doi.org/10.1002/​
tumors: comparison with conventional MRI 60 . Aoyagi T, Shuto K, Okazumi S, Shimada cncr.24604
and surgical and histopathologic findings—a H, Kazama T, Matsubara H. Apparent 68. Wang C, Zhao K, Huang Y, Ma L, Song Y,
feasibility study. American Journal of diffusion coefficient values measured Li M. Baseline FDG uptake and peripheral
Roentgenology 2009; 193: 461–70. https://doi.​ by diffusion-­weighted imaging predict lymphocyte-­monocyte ratio for predicting
org/10.2214/AJR.08.1753 chemoradiotherapeutic effect for advanced chemoradiotherapy response in patients with
52. Lee G, I H, Kim S-­J, Jeong YJ, Kim IJ, esophageal cancer. Dig Surg. 2011;28(4):252- esophageal squamous cell carcinoma. Cancer
Pak K, et al. Clinical implication of PET/ 7. Manag Res 2019; 11: 9085–93. https://doi.​
MR imaging in preoperative esophageal 61 . De Cobelli F, Giganti F, Orsenigo E, Cellina org/10.2147/CMAR.S227342
cancer staging: comparison with PET/CT, M, Esposito A, Agostini G, et al. Apparent 69. Wieder HA, Ott K, Lordick F, Becker K, Stahl
endoscopic ultrasonography, and CT. J Nucl diffusion coefficient modifications in A, Herrmann K, et al. Prediction of tumor
Med 2014; 55: 1242–47. https://doi.org/10.​ assessing gastro-­oesophageal cancer response response by FDG-­PET: comparison of the
2967/jnumed.114.138974 to neoadjuvant treatment: comparison accuracy of single and sequential studies
53 . Linder G, Korsavidou-­Hult N, Bjerner T, with tumour regression grade at histology. in patients with adenocarcinomas of the
Ahlström H, Hedberg J. 18F-­FDG-­PET/MRI European radiology. 2013;23(8):2165-74. esophagogastric junction. Eur J Nucl Med
in preoperative staging of oesophageal and 62. Lei J, Han Q, Zhu S, Shi D, Dou S, Su Z, Mol Imaging 2007; 34: 1925–32. https://doi.​
gastroesophageal junctional cancer. Clinical et al. Assessment of esophageal carcinoma org/10.1007/s00259-007-0521-3
Radiology. 2019;74(9):718-25. undergoing concurrent chemoradiotherapy 70. Bang Y-­J, Van Cutsem E, Feyereislova
54. Sharkey AR, Sah BR, Withey SJ, Bhuva with quantitative dynamic contrast-­enhanced A, Chung HC, Shen L, Sawaki A, et al.
S, Neji R, Jeljeli S, et al. Initial experience magnetic resonance imaging. Oncol Lett Trastuzumab in combination with
in staging primary oesophageal/gastro-­ 2015; 10: 3607–12. https://doi.org/10.3892/​ chemotherapy versus chemotherapy alone for
oesophageal cancer with 18F-­FDG PET/ ol.2015.3779 treatment of HER2-­positive advanced gastric
MRI. Eur J Hybrid Imaging 2021; 5(1). 63. Sun N-­N, Liu C, Ge X-­L, et al, Departments or gastro-­oesophageal junction cancer (toga):
https://doi.org/10.1186/s41824-021-00117-y of Radiology and Radiotherapy The First a phase 3, open-­label, randomised controlled
55. Matthews R, Choi M. Clinical utility of Affiliated Hospital of Nanjing Medical trial. Lancet 2010; 376: 687–97. https://doi.​
positron emission tomography magnetic University, Nanjing, China. Dynamic org/10.1016/S0140-6736(10)61121-X
resonance imaging (PET-­MRI) in contrast-­enhanced MRI for advanced 71. Lumish MA, Maron SB, Paroder V,
gastrointestinal cancers. Diagnostics esophageal cancer response assessment after Philemond S, Donoghue JAO, Schöder H,
2016; 6: 35. https://doi.org/10.3390/​ concurrent chemoradiotherapy. Diagn Interv et al. Imaging HER2-­positive metastatic
diagnostics6030035 Radiol 2018; 24: 195–202. https://doi.org/10.​ esophagogastric cancer with 89zr-­
56. Sjoquist KM, Burmeister BH, Smithers BM, 5152/dir.2018.17369 trastuzumab PET and 18F-­FDG PET. J Clin
Zalcberg JR, Simes RJ, Barbour A, et al. 64. Heethuis SE, Rossum PS, Lips IM, Goense Oncol 2021; 39: 15.
Survival after neoadjuvant chemotherapy L, Voncken FE, Reerink O, et al. Dynamic 72. Jokar N, Velikyan I, Ahmadzadehfar H,
or chemoradiotherapy for resectable contrast-­enhanced MRI for treatment Rekabpour SJ, Jafari E, Ting HH, et al.
oesophageal carcinoma: an updated meta-­ response assessment in patients with Theranostic approach in breast cancer: A
analysis. The Lancet Oncology 2011; 12: oesophageal cancer receiving neoadjuvant treasured tailor for future oncology. Clin
681–92. https://doi.org/10.1016/S1470-2045(​ chemoradiotherapy. Radiotherapy and Nucl Med 2021; 46: e410–20. https://doi.org/​
11)70142-5 Oncology: Journal of the European Society 10.1097/RLU.0000000000003678
57. Noble F, Lloyd MA, Turkington R, for Therapeutic Radiology and Oncology 73. Song IH, Lee TS, Park YS, Lee JS, Lee BC,
Griffiths E, O’Donovan M, O’Neill JR, et al. 2016; 120: 128–35. https://doi.org/10.1016/j.​ Moon BS, et al. Immuno-­PET imaging and
Multicentre cohort study to define and radonc.2016.05.009 radioimmunotherapy of 64cu-/177lu-­labeled

11 of 12 birpublications.org/bjr Br J Radiol;95:20220410
 BJR Withey et al

anti-­EGFR antibody in esophageal squamous 82. Ge X, Yu J, Wang Z, Xu Y, Pan C, Jiang L, 90. Erasmus JJ, Munden RF, Truong MT, Ho
cell carcinoma model. J Nucl Med 2016; 57: et al. Comparative study of dual energy JJ, Hofstetter WL, Macapinlac HA, et al.
1105–11. https://doi.org/10.2967/jnumed.​ CT iodine imaging and standardized Preoperative chemo-­radiation-­induced
115.167155 concentrations before and after ulceration in patients with esophageal
74. Melsens E, De Vlieghere E, Descamps B, chemoradiotherapy for esophageal cancer. cancer: a confounding factor in tumor
Vanhove C, Kersemans K, De Vos F, et al. BMC Cancer 2018; 18(1). https://doi.org/10.​ response assessment in integrated computed
Hypoxia imaging with 18f-­FAZA PET/CT 1186/s12885-018-5058-2 tomographic-­positron emission tomographic
predicts radiotherapy response in esophageal 83. Hayano K, Shuto K, Satoh A, Aoyagi T,
imaging. J Thorac Oncol 2006; 1: 478–86.
adenocarcinoma xenografts. Radiat Oncol Narushima K, Gunji H, et al. Tumor blood
https://doi.org/10.1016/S1556-0864(15)​
2018; 13: 39. https://doi.org/10.1186/s13014-​ flow change measured by CT perfusion
31614-2
018-0984-3 during chemoradiation therapy (CRT)
91. Chao Y-­K, Yeh C-­J, Lee M-­H, Wen
75. Imanishi S, Shuto K, Aoyagi T, Kono T, for monitoring response and predicting
Y-­W, Chang H-­K, Tseng C-­K, et al.
Saito H, Matsubara H. Diffusion-­weighted survival in patients with esophageal cancer.
magnetic resonance imaging for predicting Esophagus 2013; 11: 72–79. https://doi.org/​ Factors associated with false-­negative
and detecting the early response to 10.1007/s10388-013-0400-x endoscopic biopsy results after
chemoradiotherapy of advanced esophageal 84. Makari Y, Yasuda T, Doki Y, Miyata H, neoadjuvant chemoradiotherapy in
squamous cell carcinoma. Dig Surg 2013; 30: Fujiwara Y, Takiguchi S, et al. Correlation patients with esophageal squamous cell
240–48. https://doi.org/10.1159/000351435 between tumor blood flow assessed by carcinoma. Medicine (Baltimore) 2015;
76. van Rossum PSN, van Lier ALHMW, van perfusion CT and effect of neoadjuvant 94: e588. https://doi.org/10.1097/MD.​
Vulpen M, Reerink O, Lagendijk JJW, Lin therapy in advanced esophageal cancers. J 0000000000000588
SH, et al. Diffusion-­weighted magnetic Surg Oncol 2007; 96: 220–29. https://doi.org/​ 92 . Cerfolio RJ, Bryant AS, Ohja B, Bartolucci
resonance imaging for the prediction 10.1002/jso.20820 AA, Eloubeidi MA. The accuracy of
of pathologic response to neoadjuvant 85. Djuric-­Stefanovic A, Micev M, Stojanovic-­ endoscopic ultrasonography with fine-­needle
chemoradiotherapy in esophageal cancer. Rundic S, Pesko P, Saranovic D. Absolute aspiration, integrated positron emission
Radiother Oncol 2015; 115: S0167-­ CT perfusion parameter values after the tomography with computed tomography,
8140(15)00236-­4: 163–70: . https://doi.org/​ neoadjuvant chemoradiotherapy of the and computed tomography in restaging
10.1016/j.radonc.2015.04.027 squamous cell esophageal carcinoma patients with esophageal cancer after
77. Heethuis SE, Goense L, van Rossum correlate with the histopathologic tumor
neoadjuvant chemoradiotherapy. The Journal
PSN, Borggreve AS, Mook S, Voncken regression grade. European Journal of
of Thoracic and Cardiovascular Surgery.
FEM, et al. DW-­MRI and DCE-­MRI are Radiology 2015; 84: 2477–84. https://doi.org/​
2005;129(6):1232-41.
of complementary value in predicting 10.1016/j.ejrad.2015.09.025
93. Chen H, Li Y, Wu H, Sun L, Lin Q, Zhao
pathologic response to neoadjuvant 86. Cremonesi M, Garibaldi C, Timmerman R,
L, et al. 3’-deoxy-­3’-[18f]-fluorothymidine
chemoradiotherapy for esophageal cancer. Ferrari M, Ronchi S, Grana CM, et al. Interim
Acta Oncol 2018; 57: 1201–8. https://doi.org/​ 18
F-­FDG-­PET/CT during chemo-­radiotherapy pet/ct in early determination of prognosis
10.1080/0284186X.2018.1473637 in the management of oesophageal cancer in patients with esophageal squamous
78. Borggreve AS, Heethuis SE, Boekhoff MR, patients. A systematic review. Radiother Oncol cell cancer. Strahlenther Onkol 2014; 191:
Goense L, van Rossum PSN, Brosens LAA, 2017; 125: S0167-­8140(17)32589-­6: 200–212: . 141–52. https://doi.org/10.1007/s00066-014-​
et al. Optimal timing for prediction of https://doi.org/10.1016/j.radonc.2017.09.022 0744-8
pathologic complete response to neoadjuvant 87. Lordick F, Ott K, Krause B-­J, Weber WA, 94. Hotta M, Minamimoto R, Yamada K, Nohara
chemoradiotherapy with diffusion-­weighted Becker K, Stein HJ, et al. PET to assess early K, Soma D, Nakajima K, et al. Efficacy of
MRI in patients with esophageal cancer. Eur metabolic response and to guide treatment 4′-[methyl-­11C] thiothymidine PET/CT
Radiol 2020; 30: 1896–1907. https://doi.org/​ of adenocarcinoma of the oesophagogastric before and after neoadjuvant therapy for
10.1007/s00330-019-06513-0 junction: the MUNICON phase II trial. predicting therapeutic responses in patients
79. Vollenbrock SE, Voncken FEM, van Dieren Lancet Oncol 2007; 8: 797–805. https://doi.​ with esophageal cancer: a pilot study.
JM, Lambregts DMJ, Maas M, Meijer GJ, org/10.1016/S1470-2045(07)70244-9 EJNMMI Res 2019; 9(1). https://doi.org/10.​
et al. Diagnostic performance of MRI for 88. Betancourt-­Cuellar SL, Palacio DP, 1186/s13550-019-0478-9
assessment of response to neoadjuvant Benveniste MFK, Mawlawi Y, Erasmus JJ.
95. Kantarcı M, Polat P, Alper F, Eroǧlu A, Eren
chemoradiotherapy in oesophageal cancer. Pitfalls and pearls in esophageal carcinoma.
S, Okur A, et al. Comparison of CT and
Br J Surg 2019; 106: 596–605. https://doi.org/​ Semin Ultrasound CT MR 2021; 42: S0887-­
MRI for the diagnosis recurrent esophageal
10.1002/bjs.11094 2171(21)00039-­1: 535–41: . https://doi.org/​
carcinoma after operation. Dis Esophagus
80 . McCollough CH, Leng S, Yu L, Fletcher 10.1053/j.sult.2021.04.016
2004; 17: 32–37. https://doi.org/10.1111/j.​
JG. Dual- and Multi-­Energy CT: Principles, 89. Kroese TE, Goense L, van Hillegersberg
Technical Approaches, and Clinical R, de Keizer B, Mook S, Ruurda JP, et al. 1442-2050.2004.00354.x
Applications. Radiology. 2015;276(3):637-53. Detection of distant interval metastases after 96 . Shuto K, Saito H, Ohira G, Natsume T,
81. Forghani R, Mukherji SK. Advanced dual-­ neoadjuvant therapy for esophageal cancer Kono T, Tohma T, et al. Diffusion-­weighted
energy CT applications for the evaluation of with 18F-­FDG PET(/CT): a systematic MR imaging for postoperative nodal
the soft tissues of the neck. Clinical Radiology review and meta-­analysis. Dis Esophagus recurrence of esophageal squamous cell
2018; 73: 70–80. https://doi.org/10.1016/j.​ 2018; 31: 12. https://doi.org/10.1093/dote/​ cancer in comparison with FDG-­PET. Gan
crad.2017.04.002 doy055 To Kagaku Ryoho. 2009;36(12):2468-70.

12 of 12 birpublications.org/bjr Br J Radiol;95:20220410