Prognostic Value of Standardized Uptake Ratio in Patients with Trimodality Treatment of Locally Advanced Esophageal Carcinoma


Prognostic Value of Standardized Uptake Ratio in Patients
with Trimodality Treatment of Locally Advanced Esophageal
Carcinoma

Rebecca Bütof*1–3, Frank Hofheinz*4, Klaus Zöphel2,5, Julia Schmollack5, Christina Jentsch1–3, Sebastian Zschaeck1,2,
Jörg Kotzerke2,5,6, Jörg van den Hoff4,5, and Michael Baumann1–3,6–8

1Department of Radiotherapy and Radiation Oncology, Faculty of Medicine and University Hospital Carl Gustav Carus, Technische
Universität Dresden, Dresden, Germany; 2OncoRay–National Center for Radiation Research in Oncology, Dresden, Germany;
3National Center for Tumor Diseases (NCT), Partner Site Dresden, Dresden, Germany; 4PET Center, Helmholtz-Zentrum Dresden-
Rossendorf, Institute of Radiopharmaceutical Cancer Research, Dresden, Germany; 5Department of Nuclear Medicine, University
Hospital Carl Gustav Carus, Technische Universität Dresden, Dresden, Germany; 6German Cancer Consortium (DKTK), Partner Site
Dresden, Dresden, Germany; 7German Cancer Research Center (DKFZ), Heidelberg, Germany; and 8Helmholtz-Zentrum Dresden-
Rossendorf, Institute of Radiooncology–OncoRay, Dresden, Germany

The prognosis of patients with esophageal carcinoma remains
dismal despite ongoing efforts to improve treatment options. For

locally advanced tumors, several randomized trials have shown the

benefit of neoadjuvant chemoradiation followed by surgery com-

pared with surgery alone. The aim of this exploratory study was to
evaluate the prognostic value of different baseline PET parameters

and their potentially additional prognostic impact at the end of

neoadjuvant radiochemotherapy. Furthermore, the standardized
uptake ratio (SUR) as a new parameter for quantification of tumor

metabolism was compared with the conventional PET parameters

metabolically active tumor volume (MTV), total lesion glycolysis (TLG),

and SUV, taking into account known basic parameters. Methods: 18F-
FDG PET/CT was performed on 76 consecutive patients (60 ± 10 y
old, 71 men) with newly diagnosed esophageal cancer before and

during the last week of neoadjuvant radiochemotherapy. MTV of the

primary tumor was delineated with an adaptive threshold method.
The blood SUV was determined by manually delineating the aorta

in the low-dose CT scan. SURs were computed as the scan time–

corrected ratio of tumor SUVmax and mean blood SUV. Univariate

Cox regression and Kaplan–Meier analysis with respect to locore-
gional control (LRC), freedom from distant metastases (FFDM), and

overall survival (OS) was performed. Additionally, the independence

of PET parameters from standard clinical factors was analyzed with
multivariate Cox regression. Results: In multivariate analysis, 2
parameters showed a significant correlation with all endpoints:

restaging MTV and restaging SUR. Furthermore, restaging TLG

was prognostic for LRC and FFDM. For all endpoints, the largest
effect size was found for restaging SUR. The only basic factors

remaining significant in multivariate analyses were histology for OS

and FFDM and age for LRC. Conclusion: PET provides independent
prognostic information for OS, LRC, and FFDM in addition to stan-
dard clinical parameters in this patient cohort. Our results suggest

that the prognostic value of tracer uptake can be improved when

characterized by SUR rather than by SUV. Overall, our investigation

revealed a higher prognostic value for restaging parameters than

for baseline PET; therapy adjustments would still be possible at that

time. Further investigations are required to confirm these hypothesis-
generating results.

Key Words: FDG PET; SUV; SUR; MTV; prognostic value; esophageal
cancer

J Nucl Med 2019; 60:192–198
DOI: 10.2967/jnumed.117.207670

The prognosis for patients with esophageal carcinoma remains
poor despite ongoing efforts to improve treatment options. In early-

stage disease, surgical resection is the mainstay of therapy. For locally
advanced esophageal carcinoma, randomized trials have shown
the benefit of neoadjuvant radiochemotherapy followed by surgery,

compared with surgery alone (1,2).
Although this improvement is clinically highly important, the

magnitude of the effect of neoadjuvant treatment has been modest,

suggesting heterogeneity of response in individual patients. Therapy
outcome prediction based on basic clinical parameters alone is not
convincing in individual patients with esophageal carcinoma. One

potential route for predictive improvement is to combine optimized
quantitative assessment of the additional functional information

provided by 18F-FDG PET with proven basic parameters (3).
The most frequently used PET parameter for prognostic investi-

gations is the SUV. Some studies have shown that pretherapeutic

SUV has the potential to provide prognostic information in addition
to other parameters (e.g., histology, grading, T- and N-stage, and age)
in patients with esophageal carcinoma treated with neoadjuvant

radiochemotherapy followed by surgery (4). In contrast, other
studies have been published showing only a trend for significance

of SUV or even no prognostic impact on overall survival (OS)
or other clinical endpoints (5). One possible explanation for the
unsatisfactory performance of SUV so far might be the adverse

effects of well-known shortcomings of SUV quantification, espe-
cially in the clinical setting. The scan time dependence of this
parameter, interstudy variability of the arterial input function, and

susceptibility to errors in scanner calibration (6–8) confound the

Received Jan. 3, 2018; revision accepted Jun. 7, 2018.
For correspondence or reprints contact: Rebecca Bütof, Department of

Radiation Oncology University Hospital Carl Gustav Carus, Fetscherstraße
74, 01307 Dresden, Germany.
E-mail: rebecca.buetof@uniklinikum-dresden.de
*Contributed equally to this work.
Published online Aug. 30, 2018.
COPYRIGHT © 2019 by the Society of Nuclear Medicine and Molecular Imaging.

192 THE JOURNAL OF NUCLEAR MEDICINE • Vol. 60 • No. 2 • February 2019

mailto:rebecca.buetof@uniklinikum-dresden.de


reliability of the SUV as a surrogate of the metabolic uptake rate
Km of the tumor’s

18F-FDG accumulation (and thus its glucose
consumption). Therefore, the tumor-to-blood standardized uptake
ratio (SUR) has been investigated as a new, promising parameter
during the last few years. This ratio of tumor SUV to blood SUV
has been shown to correlate better with Km (derived via Patlak
analysis of dynamic PET scans) than does SUV alone (9). In addi-
tion, a reliable correction of SUR for the variability of the uptake time
is possible by converting the measured uptake values to a preselected
fixed scanning time point (10). This scan-time–normalized SUR
removes several of the shortcomings of SUV, decreased test–retest
variability (11), and provided independent prognostic informa-
tion in patients with esophageal carcinoma treated with definitive
radiochemotherapy (12).
Hence, various questions arise for patients with esophageal

carcinoma undergoing trimodality treatment. For example, what is
the prognostic value of the different baseline PET parameters,
particularly metabolically active tumor volume (MTV), which is
theoretically the most important radiobiological parameter (also
considering the difficulties of CT-based delineation in the case of
esophageal carcinoma)? Furthermore, what additional prognostic value
is to be expected of these parameters at the end of neoadjuvant
radiochemotherapy? At that point, changes in therapeutic concepts
are still possible, such as avoiding a highly risky surgery in a
patient with a poor prognosis, which would have significant clinical
consequences. Therefore, all baseline PET parameters, restaging
measurements, and fractional differences were investigated in the
present study.
A further aim of this exploratory study was to evaluate the prog-

nostic value of SUR as a new parameter for quantification of tumor
metabolism in comparison to the conventional PET parameters MTV,
total lesion glycolysis (TLG), and SUV. Finally, the determination
of specific scores and their prognostic value for therapy outcome in
patients with esophageal carcinoma receiving neoadjuvant radioche-
motherapy followed by surgery was investigated.

MATERIALS AND METHODS

Patient Characteristics

In total, 541 patients with esophageal carcinoma underwent 18F-

FDG PET/CT imaging for staging from October 2005 to December
2014 at our institution. Among them, 298 patients were treated with

palliative intent because of distant metastases or a synchronous second
cancer diagnosed by PET; 116 received curative, definitive radioche-

motherapy; and 51 had primary resection without other therapy. The
remaining 76 consecutive patients with 18F-FDG PET/CT–staged

esophageal carcinoma who received curative neoadjuvant radioche-
motherapy were retrospectively analyzed.

Evaluation of the data was approved by the Institutional Ethics Com-
mittee, and before starting treatment, all subjects provided written

informed consent to the use of their data for research.
Inclusion criteria were age greater than 18 y, histologically confirmed

esophageal carcinoma, 18F-FDG PET/CT before and at the end (during
the last week) of radiochemotherapy, no distant metastases, curative treat-

ment intention, and a minimum follow-up of 2 y.

The median age of the patients was 58 y (range, 40–80 y), and
most were male (93%). Table 1 summarizes the patient and tumor

characteristics.

Treatment

The patients underwent 3-dimensional CT-planned conformal radio-

therapy, receiving a total dose of 40 Gy applied as 2-Gy fractions, with
a dose distribution compliant with report 50 of the International

Commission on Radiation Units and Measurements (95%–107%). Gross

tumor volume was defined as primary tumor and any supect lymph nodes
on CT (short axis . 1 cm) or avid on 18F-FDG PET. The clinical
target volume was obtained by expanding this volume using a margin
of 1.5 cm (2–5 cm craniocaudally) and, after adjusting for anatomic

boundaries, adding the elective lymph node stations. Thereafter, the
volume was expanded to a planning target volume using institutional

margins of 6–7 mm.
Irradiation started concurrently with the first chemotherapy cycle.

For all patients, the chemotherapy consisted of 5-fluorouracil (in-
travenous administration of 3,000 mg/m2 over 96 h) and cisplatin

(intravenous administration of 70 mg/m2 on day 1) in weeks 1 and 4 of
the treatment. Seventy-five patients received both chemotherapy cycles;

in 1 patient, only 1 cycle was applied.

All patients underwent surgery after neoadjuvant radiochemotherapy.
In 64 patients, a complete resection was done; in 12 patients, surgery had

to be discontinued for various reasons (e.g., aortic infiltration or cardiac
arrhythmias).

18F-FDG PET/CT Protocol

All patients underwent a hybrid 18F-FDG PET/CT scan before
treatment. A second scan was performed during the last week of radio-

chemotherapy. Scans (3-dimensional PET acquisition, 3-min emission
per bed position) were performed with a Biograph 16 (Siemens Medical

Solutions Inc.).
Data acquisition started 79.8 6 22.8 min (range, 47–148 min) after

injection of 177–406 MBq of 18F-FDG. All patients had fasted for at least

6 h before 18F-FDG injection. The serum glucose concentration measured

TABLE 1
Patient and Tumor Characteristics

Characteristic Value

Age (y)

Mean ± SD 60 ± 10

Median 58

Sex

Male 71 (93)

Female 5 (7)

Histology

Squamous cell carcinoma 40 (53)

Adenocarcinoma 36 (47)

T stage

T2 9 (12)

T3 64 (84)

T4 2 (3)

Tx 1 (1)

N stage

N0 10 (13)

N1 60 (79)

N2 6 (8)

UICC stage

II 19 (25)

III 57 (75)

Data are n followed by percentage in parentheses, except for

age.

SUR IN TRIMODAL TREATED ESOPHAGEAL CANCER • Bütof et al. 193



before injection was 5.8 mmol/mL on average (range, 3.4–9.5 mmol/mL).

Tomographic images were reconstructed using attenuation-weighted
ordered-subset expectation maximization (4 iterations, 8 subsets,

and a Gaussian filter of 5 mm in full width at half maximum).

PET Data Analysis

Region-of-interest (ROI) definition and ROI analyses was per-
formed using ROVER software, version 3.0.29 (ABX). Here, we use

ROI synonymously for VOI to denote a 3-dimensional volume of
interest.

The MTV of the primary tumor was delineated in the PET data by
an automatic algorithm based on adaptive thresholding considering

the local background (13,14). The resulting delineation was inspected
visually by an experienced observer masked to patient outcome. Le-

sions were corrected manually in 5 of 76 cases with low diffuse tracer
accumulation. In 12 further cases, the primary tumor was not visible

in the second PET scan. In these cases, a small ROI (,1 mL) was
manually placed in the esophagus at the location depicted by CT

and/or based on information from endoscopy. For the delineated ROIs,
we computed SUVmax, MTV, and TLG (TLG 5 MTV · SUVmean).

The arterial blood SUV for computation of SURs was determined
as described in the supplemental material (15,16) (supplemental ma-

terials are available at http://jnm.snmjournals.org).
Lesion SURmax was computed as the uptake time–corrected ratio of

lesion SUVmax and blood SUV (BSUV). Uptake time correction to T0 5
75 min after injection was performed as described previously (10). A

value of zero for the apparent volume of distribution was assumed (i.e.,
Vr 5 0 was used in the correction formula) for the reasons discussed
elsewhere (17). The uptake time–corrected SUR is then given by

SURmax 5
T0
T

·
SUVmax
BSUV

; Eq. 1

where T is the actual time of measurement in the respective scan. In the

following, we omit the index ‘‘max’’ when denoting SUVmax/SURmax,
because maximum values of these quantities were considered throughout.

Two SUV variants were evaluated, namely normalization of injected
dose to whole-body mass and to lean body mass (18). Because the latter

variant yielded slightly superior performance (supplemental materials),

it was selected for all further statistical analysis and comparison to
other parameters. Therefore, in the following, ‘‘SUV’’ denotes lean body

mass–corrected values everywhere.
For the 4 PET parameters determined in baseline PET (MTV, TLG,

SUV, and SUR) and in restaging PET (rMTV, rTLG, rSUV, and rSUR),
their prognostic value of 3 clinical endpoints (described below) was

analyzed. Additionally, the fractional difference in these parameters
between the first and second scans was investigated. For SUV, this

difference reads

DSUV 5
SUV2 2 SUV1

SUV1
; Eq. 2

where the 1 and 2 refer to the pretherapeutic and restaging PET,

respectively. The fractional differences in the other 3 PET parameters
were computed accordingly. A summary of all investigated PET

parameters is shown in Supplemental Table 1.

Clinical Endpoints and Statistical Analysis

The 3 clinical endpoints of this study were locoregional control (LRC),
freedom from distant metastases (FFDM), and OS measured from the

start of radiotherapy to death or event. Patients who did not keep follow-
up appointments and for whom information on survival or tumor status

therefore was unavailable were censored at the date of last follow-up.
The association of OS, LRC, and FFDM with clinically relevant

parameters (age, histology, T-stage, N-stage, and UICC stage) and with

quantitative PET parameters was analyzed using univariate Cox

proportional hazards regression in which the PET parameters were
included as binarized parameters. The cutoffs used for binarization

were calculated by performing a univariate Cox regression for each
measured value. The value leading to the hazard ratio (HR) with the

highest significance was used as the cutoff. To avoid group sizes that
were too small, only values within the interquartile range were con-

sidered as potential cutoffs. The cutoffs were separately computed for
OS, LRC, and FFDM. The stability of optimal cutoffs was tested using

the bootstrap method (random resampling with replacement, 105 sam-
ples). For each sample, a univariate Cox regression was performed in

which the same cutoff as in the original data was used to define high-
and low-risk groups. Mean (sample-averaged) HR and P value were

computed. The fractions of samples yielding P values of less than 0.05
and less than 0.1, respectively, were determined. Considering parame-

ters with a mean P value of less than 0.1, the range of cutoffs for which
univariate analysis leads to a P value of less than 0.1 was determined

(supplemental materials).
The prognostic independence of PET parameters from clinically

relevant parameters was analyzed in multivariate Cox regression.

Those parameters with at least a trend for significance according to
bootstrap analysis (P , 0.1) were included. Each PET parameter was
analyzed separately together with the respective clinical parameters.
The HRs and the P values of the basic parameters were averaged over

all analyses. HRs were compared using the bootstrap method (105

samples) to determine the statistical distribution of (HR1 2 HR2),
from which the relevant P value then was derived. The added value
offered by using the PET parameters was illustrated by combined

Kaplan–Meier curves.
Statistical significance was assumed at a P value of less than 0.05.

Statistical analysis was performed with R, a language and environ-
ment for statistical computing, version 3.3.2 (19).

RESULTS

The 2-, 3-, and 5-y OS rates were 51%, 38%, and 33%, re-
spectively. These values are in line with data from the literature (20).
Overall, 63% of patients died during the observation period (last
follow-up, January 2017). The median OS was 15 mo. In our study,
LRC was 57% and FFDM was 45% among the survivors at 5 y.

Prognostic Factors for LRC

Univariate Cox regression with respect to LRC revealed a
significant correlation for age, MTV, rMTV, rTLG, ΔTLG, rSUV,
and rSUR (HR range, 2.58–4.78), whereas ΔMTV, TLG, and ΔSUR
showed a trend for significance (HR range, 1.99–2.18). Correspond-
ing Kaplan–Meier curves are shown in Supplemental Figure 1. Only
MTV, rMTV, rTLG, rSUV, and rSUR passed the stability test (Sup-
plemental Table 2).
In multivariate analysis, all included parameters were indepen-

dent prognostic factors for LRC (HR range, 2.65–4.75; Table 2).
The HR of rSUR was significantly larger than that of rSUV (P 5
0.048). No other HRs differed significantly. The best prognostic
score for this endpoint was achieved by combining age, rMTV,
and rSUR; the corresponding Kaplan–Meier curves are shown in
Figure 1.

Prognostic Factors for FFDM

None of the baseline PET parameters was prognostic for
FFDM. Only MTV showed a trend for significance in univariate
Cox regression (HR, 1.94). Nevertheless, all restaging parameters
and fractional differences were significant prognostic factors for
FFDM (HR range, 2.24–3.38). Corresponding Kaplan–Meier
curves are shown in Supplemental Figure 1. MTV did not pass

194 THE JOURNAL OF NUCLEAR MEDICINE • Vol. 60 • No. 2 • February 2019

http://jnm.snmjournals.org/


the stability test. Cutoffs of all significant parameters were stable
according to this test (Supplemental Table 2). Univariate analyses
also revealed a significant prognostic impact of histology.

On multivariate analysis, significant effects were found for
histology, rMTV, ΔMTV, rTLG, rSUV, ΔSUV, rSUR, and ΔSUR
(HR range, 2.23–5.39; Supplemental Table 3). Differences in the

HRs for rSUV and rSUR and for rSUV and
rTLG were significant (P 5 0.043 and 0.029,
respectively). No other HRs differed signifi-
cantly. Combined Kaplan–Meier curves with
respect to FFDM are shown in Figure 2.

Prognostic Factors for OS

According to univariate Cox regression,
OS was significantly associated with the
PET parameters MTV, rMTV, TLG, rTLG,
rSUV, and rSUR (HR range, 1.8–2.75).
ΔSUR showed a trend for significance
(HR, 1.85). The corresponding Kaplan–
Meier curves are shown in Supplemental
Figure 1. The stability test was passed by
MTV, rMTV, rTLG, and rSUR (Supple-
mental Table 2).
Histology was a significant prognostic

factor for OS in favor of adenocarcinomas

TABLE 2
Univariate and Multivariate Cox Regression with Respect to LRC

Univariate Multivariate

Parameter Risk HR 95%CI P Bootstrap HR 95%CI P

Clinical

Age ,62 y 3.4 1.17–9.88 0.025 NA 3.18 1.08–9.36 0.039

T stage .2 2.19 0.52–9.28 0.29 NA – – –

N stage .0 2.07 0.49–8.76 0.32 NA – – –

TNM stage .2 2.28 0.78–6.64 0.13 NA – – –

Histology SCC 1.25 0.57–2.74 0.57 NA – – –

Baseline PET

MTV .26.3 mL 3.18 1.46–6.91 0.004 0.04 2.53 1.14–5.62 0.023

TLG .121 mL 1.99 0.92–4.3 0.08 0.19 – – –

SUV .13.4 0.61 0.23–1.63 0.33 – – – –

SUR .5.66 1.99 0.75–5.29 0.17 – – – –

Restaging PET

rMTV .6.6 mL 3.59 1.58–8.14 0.002 0.024 3.64 1.59–8.34 0.002

rTLG .30.2 mL 3.53 1.59–7.83 0.002 0.025 3.65 1.63–8.17 0.002

rSUV .5.33 2.81 1.22–6.48 0.015 0.069 2.65 1.15–6.11 0.023

rSUR .3.32 4.78 1.91–11.97 <0.001 0.0082 4.75 1.9–11.9 <0.001

Fractional difference

ΔMTV .−35% 2.16 1–4.69 0.051 0.16 – – –

ΔTLG .−79.9% 2.58 1.04–6.43 0.042 0.12 – – –

ΔSUV .−38.8% 1.59 0.73–3.47 0.24 – – – –

ΔSUR .−48% 2.18 0.97–4.9 0.059 0.16 – – –

CI 5 confidence interval; NA 5 not applicable; SCC 5 squamous cell carcinoma.
In multivariate analyses, each PET parameter was analyzed separately together with age, the only significant clinical parameter

in the univariate Cox regression. HRs and P values of clinical parameters were averaged over all analyses. Bootstrap column shows

sample-averaged P value resulting from corresponding bootstrap analysis. Only PET parameters with P , 0.1 were included in
multivariate analysis.

FIGURE 1. Prognostic stratification for LRC is shown as combined Kaplan–Meier curves for age

alone (A), combination of age and baseline MTV (B), and combination of age, rMTV, and rSUR (C).

In combined curves, high risk was defined as simultaneous high risk according to all included risk

factors (Table 2 shows cutoffs). Bottom x-axis depicts time from start of radiotherapy to event or

censorship. Top x-axis depicts corresponding remaining group sizes.

SUR IN TRIMODAL TREATED ESOPHAGEAL CANCER • Bütof et al. 195



(HR, 2.58). In univariate Cox regression, UICC stage also showed
a trend for significance (HR, 2.07). These 2 parameters and the
PET parameters were included in the multivariate Cox regression.
Histology, MTV, rMTV, and rSUR were multivariate significant
prognostic factors for OS, and rTLG showed a trend for sig-
nificance (Supplemental Table 4). HRs (range, 2.06–3.19) did not
differ significantly. Combined Kaplan–Meier curves with respect
to OS are shown in Figure 3.

DISCUSSION

In this exploratory study, we investigated the prognostic value
of different baseline PET parameters and their potentially
additional value at the end of neoadjuvant radiochemotherapy in
patients with esophageal carcinoma. Moreover, the aim of this
exploratory study was to evaluate the prognostic impact of SUR as
a new parameter for quantification of tumor metabolism in
comparison to the conventional PET parameters MTV, TLG, and
SUV, taking into account known basic prognostic parameters in
this patient cohort.
Our investigation revealed several PET parameters to be

independent prognostic factors for the different clinical endpoints.
Two parameters showed a significant correlation with all endpoints

on multivariate analyses: rMTV and rSUR.
Furthermore, baseline MTV was prognostic
for OS and LRC, and rTLG was prognostic
for LRC and FFDM. The only basic clinical
factors remaining significant on multivariate
analyses performed together with the
PET parameters were histology for OS
and FFDM, in favor of adenocarcinomas,
and age for LRC.
Our findings on histology are in line

with other studies supporting better OS
rates for an adenocarcinoma subtype of
esophageal carcinoma than for squamous
cell carcinoma (21).
The fact that neither T- and N-stage nor

UICC stage was significantly correlated
with the investigated endpoints can prob-
ably be explained by the homogeneity of
our patient cohort. Most of our patients had

T3 and N1 tumors and therefore represented a comparably narrow
range of tumor burden undergoing a homogeneous treatment,
making it difficult to identify prognostic factors in the sample
reported here.
Overall, our investigation revealed a higher prognostic value for

the restaging PET parameters than for baseline PET for patient
outcome after trimodality treatment in esophageal carcinoma. This
finding is in line with a retrospective study by Mamede et al. in
which the highest prognostic value for progression-free survival
was found for body-weight–normalized rSUV (22). This finding
also agrees with a study by Lomas et al in which body-weight–
normalized rSUV was the only remaining prognostic factor for OS
on multivariate analysis (23).
Furthermore, the restaging PET parameters alone have signif-

icantly higher prognostic impact than the corresponding relative
changes in the respective parameters between the 2 PET
examinations. For a prognostic score on patient outcome, the
PET-based volume parameter rMTV was combined with the
uptake parameter rSUR and histology. This score significantly
improved outcome prediction for patients with esophageal carci-
noma treated with neoadjuvant radiochemotherapy followed by
surgery for all investigated clinical endpoints, with the largest
effect being for LRC (Figs. 1–3). By determination of such a

prognostic score, a therapy adjustment
would be feasible since within our protocol
the second PET/CT is performed at a time
when changes in therapeutic concepts are
still possible—for example, avoidance of
risky surgery in patients with a poor progno-
sis, or intensification of chemotherapy in pa-
tient with a high risk for metastatic spread.
In our study, baseline SUV was not

prognostic for any of the investigated
endpoints (independent of weight normal-
ization; supplemental materials), as agrees
with most published investigations in sim-
ilar patient groups (22,24,25). Also, the use
of SUR instead of SUV as the baseline
parameter did not improve the prognostic
value of uptake parameters in our patient
cohort, as agrees with a recent publication
by Arnett et al. that found no significant

FIGURE 2. Prognostic stratification for FFDM is shown as combined Kaplan–Meier curves for

histology alone (A), combination of histology and rMTV (B), and combination of histology and

rSUR (C). In combined curves, high risk was defined as simultaneous high risk according to all

included risk factors (Supplemental Table 3 shows cutoffs). Bottom x-axis depicts time from start

of radiotherapy to event or censorship. Top x-axis depicts corresponding remaining group sizes.

FIGURE 3. Prognostic stratification for OS is shown as combined Kaplan–Meier curves for

histology alone (A), combination of histology and baseline MTV (B), and combination of histology,

rMTV, and rSUR (C). In combined curves, high risk was defined as simultaneous high risk according

to all included risk factors (Supplemental Table 4 shows cutoffs). Bottom x-axis depicts time from

start of radiotherapy to event or censorship. Top x-axis depicts corresponding remaining group sizes.

196 THE JOURNAL OF NUCLEAR MEDICINE • Vol. 60 • No. 2 • February 2019



correlation between baseline SUR and OS or histopathologic re-
sponse (24). This suggests that pretherapeutic uptake metabolism
per se is not prognostic or has only a low effect in patients with
esophageal carcinoma receiving neoadjuvant radiochemotherapy.
For ΔSUV, we did not find any significant effect on OS, as agrees
with some studies (22,24,26–28) but disagrees with others (20,29–
31). Also, the use of ΔSUR apparently did not improve the cor-
relation with our clinical endpoints. Only for FFDM—not for OS
and LRC—we found a significant effect in our study. Regarding
OS, this finding agrees with the results of Arnett et al. (24). Taken
together, these ambiguous results might indicate a limited prog-
nostic value or clinically irrelevant effect of changes in uptake
metabolism in this patient cohort.
For all endpoints, the largest effect size (HR) of all PET

parameters was found for rSUR, with the HR difference between
rSUR and rSUV, respectively, being significant for LRC and
FFDM, which further indicates that SUR has the potential to
increase the prognostic value compared with SUV. In contrast, the
HRs for rSUR and volume-based parameters (MTV, TLG) did not
differ significantly. Therefore, the question of whether rSUR is
also superior to these parameters cannot be finally answered on the
basis of our results. It is important to note that for each parameter,
the optimal cutoff was used for the analysis, resulting in an explorative
analysis. The robustness of these cutoffs must be tested in a validation
cohort and finally confirmed in a prospective multicenter trial.
As to why SUR outperforms SUV, it is relevant to note that,

contrary to SUV, SUR is (approximately) proportional to the
tumors’ 18F-FDG uptake rate, Km, across different investigations,
even in the presence of interindividual uptake time and BSUV
variability (9,10,17). Assuming glucose metabolism is prognostic
of treatment outcome, SUR can be expected to have a higher
prognostic value than SUV. Although uptake time correction of
SUV is possible (10), it suffers from specific uncertainties and
does not suffice in the present study to improve the performance
of SUV (supplemental materials). Omitting uptake time correction
from SUR computation (and thus reducing it to SUV normaliza-
tion to BSUV) is not an option because uptake time correlates with
BSUV (BSUV decreases over time). It is only the combined cor-
rection which makes SUR proportional to Km to a rather good
approximation. Notwithstanding, it is conceivable that part of
the improvement achieved by SUR will be related not to compen-
sation of true intersubject BSUV variability but to elimination of
the usual factors adversely affecting SUV quantitation (scanner cali-
bration, dose, and weight uncertainties, for example). SUR shares this
advantage with other ratio methods (e.g., the tumor-to-liver ratio) but
offers further benefits by eliminating the influence of physiologic
BSUV and uptake time variability. In our view, existing data indicate
that intersubject BSUV variability is, in fact, much larger than what
could be expected from inaccurate or erroneous SUV determination
alone, but ultimate proof of this conjecture would require comparing
image-derived BSUVs with blood samples in a dedicated prospective
study. In any case, for practical application of SUR, it is not necessary
to differentiate between true and spurious BSUV variability because
both are simultaneously corrected (ensuring the said proportionality
of SUR to Km).
Furthermore, our results demonstrate that the correlation of

baseline MTV with LRC and OS, and of rMTV with all clinical
endpoints, is significant. Other recent studies also suggest MTV to
have prognostic value regarding OS and recurrence-free survival,
for example (32). MTV has also been shown to correlate with
pathologic response after neoadjuvant chemoradiation as a

predictive marker for patient outcome (33), as is in line with
the results reported here. From a radiobiologic point of view,
tumor volume as a surrogate of the number of cancer stem cells
is highly relevant because the number of cancer stem cells in a
tumor correlated with both likelihood of local control and like-
lihood of metastases (34). The results of our study are in line
with these expectations. Thus, on the basis of this investigation,
the importance of tumor volume (or its metabolic active part) to
patient outcome can be demonstrated in a further entity and
extends previously published data on, for example, lung cancer
or head and neck carcinomas (35–37). Because MTV and SUR are
independent prognostic parameters in this study, it may be specu-
lated that both the overall number of cancer stem cells and tumor
metabolism/proliferation may play a major role in therapy outcome
for this patient cohort.
It should be noted that in the present study, MTV was determined

using an essentially fully automated delineation algorithm. Although
several viable automated algorithms have been published (38–43), in
many institutions MTV is still determined by manual delineation or
by application of a fixed absolute or relative threshold, which suffers
from well-known limitations such as intra- and interobserver vari-
ability as well as a possibly background-dependent bias. Therefore,
the prognostic value of MTV may potentially be less convincing
when the lesions are delineated manually in the usual fashion. In
contrast, SURmax has the clear advantage of being quite insensitive
to the details of tumor delineation (unambiguous determination
of lesion SUVmax). SUR also is quite insensitive to details of
the aorta ROI definition (used for computation of blood SUV),
since it can be defined reliably in the attenuation CT scan (Sup-
plemental Fig. 2). Residual ambiguity or variability of ROI position
and delineation does not cause notable changes in derived blood SUV,
because the applied substantial safety margin of 8 mm ensures that
partial-volume effects do not become relevant.
The analysis of all baseline parameters revealed a combination

of age and MTV to have the highest impact for a prognostic score
in this patient cohort, but there was hardly any additional value
over the prognostic power of the individual parameters alone
(Figs. 1–3).
If the prognostic value can be confirmed in a validation cohort,

the predictive value of the described PET parameters and com-
bined scores needs to be addressed in prospective stratification
studies or intervention studies.

CONCLUSION

PET provides independent prognostic information for OS, LRC,
and FFDM in addition to standard clinical parameters in patients
with esophageal carcinoma treated with neoadjuvant radiochemo-
therapy followed by surgery. Among the investigated uptake-based
parameters, SUV and SUR determined in the restaging PET were
independent prognostic factors for all investigated clinical end-
points. In all cases, the effect size was larger for SUR than for SUV.
These results suggest that the prognostic value of tracer uptake can
be improved when characterized by SUR rather than by SUV.
Furthermore, a significant impact of MTV before and after

neoadjuvant radiochemotherapy could be shown. Thus, the radio-
biologic importance of the tumor volume for patient outcome was
demonstrated in a further entity and adds to previously published
data on other tumors.
Overall, our investigation revealed the restaging PET parame-

ters to be more prognostic than the baseline PET parameters for

SUR IN TRIMODAL TREATED ESOPHAGEAL CANCER • Bütof et al. 197



patient outcome after trimodality treatment of esophageal carcinoma.
One important fact is that within our protocol, therapy adjustments
would still be possible at that point in time. Further investigations
are required to confirm these hypothesis-generating results.

DISCLOSURE

This work was supported in parts by the German Federal
Ministry of Education and Research (BMBF contract 03ZIK041).
No other potential conflict of interest relevant to this article was
reported.

REFERENCES

1. Stahl M, Walz MK, Stuschke M, et al. Phase III comparison of preoperative

chemotherapy compared with chemoradiotherapy in patients with locally ad-

vanced adenocarcinoma of the esophagogastric junction. J Clin Oncol. 2009;27:

851–856.

2. van Hagen P, Hulshof MC, van Lanschot JJ, et al. Preoperative chemoradiother-

apy for esophageal or junctional cancer. N Engl J Med. 2012;366:2074–2084.

3. Baumann M, Krause M, Overgaard J, et al. Radiation oncology in the era of

precision medicine. Nat Rev Cancer. 2016;16:234–249.

4. Omloo JM, van Heijl M, Hoekstra OS, van Berge Henegouwen MI, van

Lanschot JJ, Sloof GW. FDG-PET parameters as prognostic factor in esophageal

cancer patients: a review. Ann Surg Oncol. 2011;18:3338–3352.

5. Van de Wiele C, Kruse V, Smeets P, Sathekge M, Maes A. Predictive and

prognostic value of metabolic tumour volume and total lesion glycolysis in solid

tumours. Eur J Nucl Med Mol Imaging. 2013;40:290–301.

6. Hamberg LM, Hunter GJ, Alpert NM, Choi NC, Babich JW, Fischman AJ. The

dose uptake ratio as an index of glucose metabolism: useful parameter or over-

simplification? J Nucl Med. 1994;35:1308–1312.

7. Huang SC. Anatomy of SUV: standardized uptake value. Nucl Med Biol. 2000;

27:643–646.

8. Keyes JW Jr. SUV: standard uptake or silly useless value? J Nucl Med. 1995;

36:1836–1839.

9. van den Hoff J, Oehme L, Schramm G, et al. The PET-derived tumor-to-blood

standard uptake ratio (SUR) is superior to tumor SUV as a surrogate parameter

of the metabolic rate of FDG. EJNMMI Res. 2013;3:77.

10. van den Hoff J, Lougovski A, Schramm G, et al. Correction of scan time depen-

dence of standard uptake values in oncological PET. EJNMMI Res. 2014;4:18.

11. Hofheinz F, Apostolova I, Oehme L, Kotzerke J, van den Hoff J. Test-retest

variability in lesion SUV and lesion SUR in 18F-FDG PET: an analysis of data

from two prospective multicenter trials. J Nucl Med. 2017;58:1770–1775.

12. Bütof R, Hofheinz F, Zophel K, et al. Prognostic value of pretherapeutic tumor-

to-blood standardized uptake ratio in patients with esophageal carcinoma. J Nucl

Med. 2015;56:1150–1156.

13. Hofheinz F, Langner J, Petr J, et al. An automatic method for accurate volume

delineation of heterogeneous tumors in PET. Med Phys. 2013;40:082503.

14. Hofheinz F, Potzsch C, Oehme L, et al. Automatic volume delineation in onco-

logical PET: evaluation of a dedicated software tool and comparison with manual

delineation in clinical data sets. Nuklearmedizin. 2012;51:9–16.

15. Patlak C, Blasberg R, Fenstermacher J. Graphical evaluation of blood-to-brain

transfer constants from multiple-time uptake data. J Cereb Blood Flow Metab.

1983;3:1–7.

16. Patlak C, Blasberg R. Graphical evaluation of blood-to-brain transfer constants

from multiple-time uptake data: generalizations. J Cereb Blood Flow Metab.

1985;5:584–590.

17. Hofheinz F, Hoff J, Steffen IG, et al. Comparative evaluation of SUV, tumor-to-

blood standard uptake ratio (SUR), and dual time point measurements for as-

sessment of the metabolic uptake rate in FDG PET. EJNMMI Res. 2016;6:53.

18. Sugawara Y, Zasadny KR, Neuhoff AW, Wahl RL. Reevaluation of the standard-

ized uptake value for FDG: variations with body weight and methods for cor-

rection. Radiology. 1999;213:521–525.

19. RCoreTeam. R: A Language and Environment for Statistical Computing. Vienna,

Austria: R Foundation for Statistical Computing; 2014.

20. Cuenca X, Hennequin C, Hindie E, et al. Evaluation of early response to con-

comitant chemoradiotherapy by interim 18F-FDG PET/CT imaging in patients

with locally advanced oesophageal carcinomas. Eur J Nucl Med Mol Imaging.

2013;40:477–485.

21. Eil R, Diggs BS, Wang SJ, Dolan JP, Hunter JG, Thomas CR. Nomogram for

predicting the benefit of neoadjuvant chemoradiotherapy for patients with esoph-

ageal cancer: a SEER-Medicare analysis. Cancer. 2014;120:492–498.

22. Mamede M, Abreu ELP, Oliva MR, Nose V, Mamon H, Gerbaudo VH. FDG-

PET/CT tumor segmentation-derived indices of metabolic activity to assess

response to neoadjuvant therapy and progression-free survival in esophageal

cancer: correlation with histopathology results. Am J Clin Oncol. 2007;30:

377–388.

23. Lomas H, Hoffe S, Weber J. Post chemoradiation PET SUV is highly pre-

dictive of overall survival in esophageal cancer. J Nucl Med Radiat Ther.

2012;3:125.

24. Arnett AL, Merrell KW, Macintosh EM, et al. Utility of 18F-FDG PET for

predicting histopathologic response in esophageal carcinoma following chemo-

radiation. J Thorac Oncol. 2017;12:121–128.

25. Metser U, Rashidi F, Moshonov H, et al. 18F-FDG-PET/CT in assessing response

to neoadjuvant chemoradiotherapy for potentially resectable locally advanced

esophageal cancer. Ann Nucl Med. 2014;28:295–303.

26. Kim MK, Ryu JS, Kim SB, et al. Value of complete metabolic response by 18F-

fluorodeoxyglucose-positron emission tomography in oesophageal cancer for

prediction of pathologic response and survival after preoperative chemoradio-

therapy. Eur J Cancer. 2007;43:1385–1391.

27. Malik V, Lucey JA, Duffy GJ, et al. Early repeated 18F-FDG PET scans during

neoadjuvant chemoradiation fail to predict histopathologic response or sur-

vival benefit in adenocarcinoma of the esophagus. J Nucl Med. 2010;51:1863–

1869.

28. Tamandl D, Gore RM, Fueger B, et al. Change in volume parameters induced by

neoadjuvant chemotherapy provide accurate prediction of overall survival after

resection in patients with oesophageal cancer. Eur Radiol. 2016;26:311–321.

29. Brücher BL, Weber W, Bauer M, et al. Neoadjuvant therapy of esophageal

squamous cell carcinoma: response evaluation by positron emission tomography.

Ann Surg. 2001;233:300–309.

30. Huang JW, Yeh HL, Hsu CP, et al. To evaluate the treatment response of locally

advanced esophageal cancer after preoperative chemoradiotherapy by FDG-PET/CT

scan. J Chin Med Assoc. 2015;78:229–234.

31. Myslivecek M, Neoral C, Vrba R, et al. The value of 18F-FDG PET/CT in

assessment of metabolic response in esophageal cancer for prediction of histo-

pathological response and survival after preoperative chemoradiotherapy.

Biomed Pap Med Fac Univ Palacky Olomouc Czech Repub. 2012;156:171–179.

32. Chang S, Kim SJ. Prediction of recurrence and mortality of locally advanced

esophageal cancer patients using pretreatment F-18 FDG PET/CT parameters:

intratumoral heterogeneity, SUV, and volumetric parameters. Cancer Biother

Radiopharm. 2016;31:1–6.

33. Blom RL, Steenbakkers IR, Lammering G, et al. PET/CT-based metabolic tu-

mour volume for response prediction of neoadjuvant chemoradiotherapy in oeso-

phageal carcinoma. Eur J Nucl Med Mol Imaging. 2013;40:1500–1506.

34. Baumann M, Krause M, Hill R. Exploring the role of cancer stem cells in

radioresistance. Nat Rev Cancer. 2008;8:545–554.

35. Bütof R, Dubrovska A, Baumann M. Clinical perspectives of cancer stem cell

research in radiation oncology. Radiother Oncol. 2013;108:388–396.

36. Linge A, Lohaus F, Lock S, et al. HPV status, cancer stem cell marker expres-

sion, hypoxia gene signatures and tumour volume identify good prognosis sub-

groups in patients with HNSCC after primary radiochemotherapy: a multicentre

retrospective study of the German Cancer Consortium Radiation Oncology

Group (DKTK-ROG). Radiother Oncol. 2016;121:364–373.

37. Soliman M, Yaromina A, Appold S, et al. GTV differentially impacts locore-

gional control of non-small cell lung cancer (NSCLC) after different fraction-

ation schedules: subgroup analysis of the prospective randomized CHARTWEL

trial. Radiother Oncol. 2013;106:299–304.

38. Black QC, Grills IS, Kestin LL, et al. Defining a radiotherapy target with pos-

itron emission tomography. Int J Radiat Oncol Biol Phys. 2004;60:1272–1282.

39. Boellaard R, Krak NC, Hoekstra OS, Lammertsma AA. Effects of noise, image

resolution, and ROI definition on the accuracy of standard uptake values: a

simulation study. J Nucl Med. 2004;45:1519–1527.

40. Erdi YE, Mawlawi O, Larson SM, et al. Segmentation of lung lesion volume by

adaptive positron emission tomography image thresholding. Cancer. 1997;80:

2505–2509.

41. Nestle U, Kremp S, Schaefer-Schuler A, et al. Comparison of different methods

for delineation of 18F-FDG PET-positive tissue for target volume definition in

radiotherapy of patients with non-small cell lung cancer. J Nucl Med. 2005;46:

1342–1348.

42. van Dalen JA, Hoffmann AL, Dicken V, et al. A novel iterative method for lesion

delineation and volumetric quantification with FDG PET. Nucl Med Commun.

2007;28:485–493.

43. Vauclin S, Doyeux K, Hapdey S, Edet-Sanson A, Vera P, Gardin I. Development

of a generic thresholding algorithm for the delineation of 18FDG-PET-positive

tissue: application to the comparison of three thresholding models. Phys Med

Biol. 2009;54:6901–6916.

198 THE JOURNAL OF NUCLEAR MEDICINE • Vol. 60 • No. 2 • February 2019