ORIGINAL ARTICLE BURGIO ET AL.

Correlation of Tumor Response on

Computed Tomography With
Pathological Necrosis in Hepatocellular
Carcinoma Treated by
Chemoembolization Before Liver
Transplantation
Marco Dioguardi Burgio,1 Maxime Ronot,1,5,6 Onorina Bruno,1 Claire Francoz,2
Valerie Paradis,3,5,6 Laurent Castera,2 François Durand,2,5 Olivier Soubrane,4-6 and
Valerie Vilgrain1,5,6
Departments of 1Radiology, 2Hepatology, 3Pathology, and 4Surgery, University Hospitals Paris Nord Val de Seine, Beaujon, Cli-
chy, Hauts-de-Seine, France; 5University Paris Diderot, Sorbonne Paris Cite, Paris, France; and 6INSERM U1149, Centre de
Recherche Biomedicale Bichat-Beaujon, CRB3, Paris, France

The purpose of this article was to compare the results of Response Evaluation Criteria in Solid Tumors (RECIST), modi-
fied Response Evaluation Criteria in Solid Tumors (mRECIST), and European Association for the Study of the Liver
(EASL) criteria for the evaluation of tumor necrosis in patients treated with transarterial chemoembolization before liver
transplantation (LT) for hepatocellular carcinoma. Response to treatment was evaluated on computed tomography scan by
2 independent readers based on RECIST, mRECIST, and EASL criteria, and compared with tumor necrosis assessed by
explant pathology. Necrosis was defined as major when >90%. Factors associated with major necrosis were tested by multi-
variate analysis. Fifty-eight patients (53 males; mean age, 54 years; range, 31-64 years) were included with 88 nodules.
Fifty-one (58%) nodules were shown to have major necrosis. Among them readers 1 and 2 identified a complete response
(CR) according to RECIST, mRECIST, and EASL criteria in 2 (4%), 47 (92%), and 47 (92%), and 1 (2%), 45 (88%),
and 45 (88%) nodules, respectively. However, 12-14 of 59 nodules classified as CR on mRECIST or EASL criteria were
found to have intermediate or minor necrosis (overestimation in 20%-24% of the patients). Combining the classification of
CR by mRECIST and EASL criteria and complete lipiodol deposition reduced the overestimation to 11%. Among 59
nodules classified with a CR according to mRECIST or EASL, those with complete lipiodol deposition (n 5 36, 61%)
had a higher rate of necrosis than those with incomplete lipiodol deposition (n 5 23, 39%): 95% versus 68% and 95% ver-
sus 63% for reader 1 and 2, respectively. In conclusion, CR based on mRECIST/EASL combined with complete lipiodol
deposition was better for identification of major tumor necrosis. Even in the presence of CR according to mRECIST/
EASL, incomplete lipiodol deposition should be considered indicative of substantial viable tumor remnant.

Liver Transplantation 22 1491–1500 2016 AASLD.

Received March 18, 2016; accepted July 26, 2016.

Hepatocellular carcinoma (HCC) represents the fifth

Abbreviations: %nec, mean percentage of tumor necrosis upon patho-
logic analysis; CI, confidence interval; CR, complete response; CT,
computed tomography; EASL, European Association for the Study of
the Liver; HCC, hepatocellular carcinoma; HCV, hepatitis C virus;
LT, liver transplantation; mRECIST, modified Response Evaluation
Criteria in Solid Tumors; OR, odds ratio; PD, progressive disease;
PR, partial response; RECIST, Response Evaluation Criteria in
Solid Tumors; SD, stable disease; TACE, transarterial
chemoembolization.
most common cancer worldwide.(1) Current recom-
mendations on the management of patients with HCC
are based on various elements including tumor charac-
teristics (number, size of lesions, vascular invasion,
extrahepatic disease), liver function reserve, and clinical
status.(2,3) Transarterial chemoembolization (TACE)
is the reference treatment in patients with intermediate
disease (ie, not eligible for curative treatment such as

ORIGINAL ARTICLE | 1491


BURGIO ET AL.
transplantation, surgical resection, or radiofrequency
ablation) and sufficiently preserved liver function.(2,3)
Otherwise, TACE is recommended (1) to prevent
tumor progression to outside acceptable criteria requir-
ing removal of candidates from the transplantation
waiting list, (2) to test tumor biology by identifying
patients with a high or low probability of cancer pro-
gression, (3) to decrease the recurrence of HCC after
liver transplantation (LT) because patients with com-
plete or nearly complete tumor necrosis have better
overall survival,(4,5) and (4) to help balance the priority
of HCC and non-HCC candidates for LT.(6,7)
The aim of TACE is to induce tumor necrosis,
which is first assessed on imaging and later by patho-
logic examination of the liver explant. Although
Response Evaluation Criteria in Solid Tumors
(RECIST)(8) is still used in LT centers, this classifica-
tion has been shown to markedly underestimate tumor
necrosis following locoregional treatment.(9) Indeed
RECIST evaluates tumor response according to
changes in tumor size and does not take into account
the tumor necrosis induced by treatment.
Thus, both the European Association for the Study
of the Liver (EASL) and the American Association for
the Study of Liver Diseases have proposed new imag-
ing criteria referred to as the EASL(10) and modified
Response Evaluation Criteria in Solid Tumors
(mRECIST)(11) criteria, respectively, which were
designed to improve evaluation of tumor response.
Unlike RECIST 1.1, the EASL and mRECIST clas-
sifications of tumor response are based on the analysis
of the viable or hypervascular portions of the tumor on
contrast-enhanced imaging.
Overall response according to EASL or mRECIST
has been shown to be an independent positive prog-
nostic factor and associated with improved survival in
Address reprint requests to Maxime Ronot, M.D., Ph.D., Depart-
ment of Radiology, University Hospitals Paris Nord Val de Seine,
Beaujon, 100 bd du General Leclerc, 92110 Clichy, Hauts-de-Seine,
France. Telephone: (33)-1-40-87-58-95; FAX: (33)-1-40-87-17-
24; E-mail: maxime.ronot@bjn.aphp.fr
Additional supporting information may be found in the online ver-
sion of this article.
Copyright V
C 2016 by the American Association for the Study of Liver
Diseases.
View this article online at wileyonlinelibrary.com.
DOI 10.1002/lt.24615
Potential conflict of interest: Nothing to report.
1492 | ORIGINAL ARTICLE
15276473, 2016, 11, Downloaded from https://aasldpubs.onlinelibrary.wiley.com/doi/10.1002/lt.24615 by Universita Di Firenze Sistema, Wiley Online Library on [24/07/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
LIVER TRANSPLANTATION, November 2016
patients treated with TACE before transplantation.(12-
15)
However, there are very few comparisons of tumor
assessment based on these imaging criteria and the
pathological examination of treated HCC. In a recent
study, Bargellini et al.(16) showed that mRECIST
could overestimate the amount of tumor necrosis but is
accurate and reproducible. However, these authors did
not compare their results to the other imaging criteria
such as RECIST 1.1 or EASL.
Thus, the aim of this study was to compare tumor
response using RECIST 1.1, mRECIST, and EASL
assessed on computed tomography (CT) with the
pathological assessment of tumor necrosis, in a con-
secutive group of patients who underwent LT for
HCC.
Patients and Methods
PATIENT AND TUMOR
SELECTION
This retrospective study was performed at the Depart-
ment of Radiology, University Hospital of Beaujon.
Our study was performed with approval of the institu-
tional review board.
All patients were identified who were treated by
LT for HCC between 2010 and 2014 and who
underwent at least 1 session of TACE before LT. LT
was indicated for HCC based on the Milan criteria.
Inclusion criteria were (1) the presence of at least 1
HCC according to EASL–European Organisation
for Research and Treatment of Cancer clinical prac-
tice guidelines,(3) (2) at least 1 session of TACE
before LT, (3) contrast-enhanced CT before the
TACE procedure and between TACE and LT.
Lesions from selected patients were excluded when
(1) a definite correlation between CT images and the
pathological specimen was not possible or (2) HCC
was treated after TACE by another locoregional
treatment (eg, ablation).
A total of 528 patients underwent LT during the
study period. Ninety-two (17%) of them underwent
TACE. Eighteen (19%) of these were excluded due
to incomplete imaging workup, 11 (12%) because the
radiopathological correlation was not feasible, and 5
(6%) patients were treated with another locoregional
treatment after TACE. The final population of
this study included 58 patients (53 males, 91%) with
88 HCC nodules. Figure 1 shows the patient
flowchart.

LIVER TRANSPLANTATION, Vol. 22, No. 11, 2016

FIG. 1. Flowchart for the study population.

TACE PROCEDURE
A multidisciplinary team decided all TACE procedures,
and imaging results were reviewed before the TACE pro-
cedure by the interventional radiology team to adapt treat-
ment to tumor characteristics. TACE procedures were
performed under local anesthesia by senior interventional
radiologists. Diagnostic visceral arteriography was first
performed by digital subtraction angiography or 3-
dimensional cone beam CT to identify tumor arterial sup-
ply. Treatment was as selective as possible and was per-
formed with a 2.7-Fr or 2.4-Fr microcatheter depending
on tumor distribution. TACE was classified as selective if
only the tumoral feeders were involved and as nonselective
in the other cases. TACE was conventional in all patients
and included an intraarterial injection of a mixture of che-
motherapy (75 mg of doxorubicin; Adriamycin; Pharma-
cia Upjohn, Kalamazoo, MI), emulsified in poppy seed oil
(Lipiodol, Gerbet, France). Embolization was performed
by injection of gelatin sponge (Gelitaspon, Gelita Medical
B.V., Amsterdam, the Netherlands) or polyvinyl alcohol
particles (Bead Block, Biocompatibles, Farnham, UK)
until near stasis. The delay between treatment and LT
and the number of sessions per patient and per lesion
were evaluated in each patient.
IMAGING
CT Protocol
Abdominal CT was performed 4-6 weeks after TACE
with multidetector CT (64-detector LightSpeed VCT;
15276473, 2016, 11, Downloaded from https://aasldpubs.onlinelibrary.wiley.com/doi/10.1002/lt.24615 by Universita Di Firenze Sistema, Wiley Online Library on [24/07/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
BURGIO ET AL.
GE Healthcare, Milwaukee, WI). After an unen-
hanced abdominal scan, a nonionic iodinated contrast
agent containing 350 mg I mL–1 was administered
intravenously through a 16-18 gauge catheter. A mean
2 mL/kg of contrast agent was injected via an antecu-
bital vein at 4 mL/s. No oral contrast medium was giv-
en. Arterial, portal, and late venous phase acquisitions
were obtained 35, 80, and 180 seconds respectively,
following administration of contrast medium (2.5-mm
maximum thickness).
Image Analysis and Tumor Response
Evaluation
Pre- and post-TACE CT were reviewed by 2 radiol-
ogists (M.D.B. and M.R. with 4 and 8 years of experi-
ence, respectively, in the field of abdominal imaging
and liver oncology) using the Picture Archiving and
Communication System. Each radiologist indepen-
dently read all CT scans and was blinded to patients’
clinical and pathological data. All images were shown
to radiologists in random order during 2 separate read-
ing sessions.
Pre- and Post-TACE Features Analysis
The following features were recorded on the pre-
TACE CT: number, maximum diameter in an axial
section,(8) and location of HCCs. The presence of a
peritumoral capsule was noted for each tumor. The
presence of a nontumoral thrombus defined as com-
plete or partial lack of enhancement of the portal vein
and/or of its branches on the portal venous phase with
no arterial phase endovascular tumor enhancement was
also recorded. At baseline, lesions were considered
when hypervascular. Tumors with no intratumoral
arterial enhancement were not included in the
analysis.(8,10,11)
The following features were noted for each nodule
on the post-TACE CT: maximum diameter, maxi-
mum unidimensional-enhanced diameter on arterial
phase images, the product of the axial arterial phase
bidimensional-enhanced diameters, and the presence
of lipiodol deposition. Lipiodol deposition was classi-
fied as complete when it covered the entire nodule and
incomplete if it did not.
Response to Treatment
Before the study, readers tested criteria in a training set
of 10 patients who were not included in the study. A
debriefing session was held to discuss any difficulties,
ORIGINAL ARTICLE | 1493
 15276473, 2016, 11, Downloaded from https://aasldpubs.onlinelibrary.wiley.com/doi/10.1002/lt.24615 by Universita Di Firenze Sistema, Wiley Online Library on [24/07/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
BURGIO ET AL. LIVER TRANSPLANTATION, November 2016



FIG. 2. Examples of HCC showing CR after TACE according to mRECIST and EASL criteria but with different lipiodol deposi-
tion. The upper line (A and B) corresponds to a HCC developed in segment 7 in a 65-year-old male with HCV-related cirrhosis. On
contrast-enhanced CT at arterial phase before TACE, (A) the lesion was ill defined. After 1 session of conventional TACE, (B) the
lesion showed no residual contrast uptake on arterial phase images corresponding to a CR according to both the mRECIST and
EASL criteria, with incomplete lipiodol deposition. Pathological examination showed 63% of tumor necrosis. The bottom line (C and
D) corresponds to a HCC developed in segment 5 in a 62-year-old male with alcohol-related cirrhosis. On contrast-enhanced CT at
arterial phase before TACE, (C) the lesion showed marked contrast uptake. After 1 session of conventional TACE, (D) the lesion
showed no residual contrast uptake on arterial phase images corresponding to a CR according to both the mRECIST and EASL cri-
teria, with complete lipiodol deposition. Pathological examination showed 100% of tumor necrosis.


to establish a standard protocol, and to select slice lev-
els. Each radiologist then read all CT scans indepen-
dently and was blinded to patients’ clinical data and
outcome.
Response to treatment was evaluated by the readers
on the last CT before LT according to RECIST 1.1,
mRECIST, and EASL criteria.(8,10,11) The mean delay
between the last CT and LT was 50 6 32 days, and the
mean delay between the last TACE and LT was
6.2 6 2.9 months. Tumor response was analyzed for all
criteria on a nodule-by-nodule basis (Table 1).
Evaluation by RECIST 1.1 criteria was performed
using the largest diameter of each lesion on an axial
slice. mRECIST and EASL criteria used 1-
(mRECIST) or bi-dimensional (EASL) axial images of
the viable portion of each lesion defined as the enhanced
portion on arterial phase contrast-enhanced CT images.
Pathological Technique and Analysis
The study coordinator matched the CT images and
the pathological specimen in consensus with the

1494 | ORIGINAL ARTICLE

 15276473, 2016, 11, Downloaded from https://aasldpubs.onlinelibrary.wiley.com/doi/10.1002/lt.24615 by Universita Di Firenze Sistema, Wiley Online Library on [24/07/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
LIVER TRANSPLANTATION, Vol. 22, No. 11, 2016 BURGIO ET AL.

TABLE 1. Tumor Response According to RECIST 1.1, backward selection). Interreader agreement for the dif-
mRECIST, and EASL Criteria
ferent imaging criteria of tumor response (ie, RECIST
NODULE (n 5 88)
1.1, mRECIST, and EASL) was evaluated using kap-
RECIST 1.1 mRECIST EASL pa statistics. Kappa coefficients between 0.00 and 0.20
Reader 1 indicated slight agreement; 0.21 and 0.40, fair agree-
CR 2 (2%) 59 (67%) 59 (37%) ment; 0.41 and 0.60, moderate agreement; 0.61 and
%nec 100% 82% 82%
PR 24 (27%) 12 (14%) 9 (10%) 0.80, substantial agreement; and 0.81 and 1.00, almost
%nec 80% 42% 41% perfect agreement. P values  0.05 were considered to
SD 53 (60%) 13 (15%) 16 (18%) be significant. All analyses were performed with the
%nec 62% 20% 25%
PD 9 (10%) 4 (4%) 4 (4%) Statistical Package for the Social Sciences software,
%nec 32% 7% 7% version 20.0 (SPSS Inc., Chicago, IL).
Reader 2
CR 1 (1%) 59 (67%) 59 (67%)

PR
%nec 100%
33 (37%)
80%
13 (15%)
80%
10 (11%)
%nec 72% 56% 63%
SD 49 (55%) 12 (14%) 15 (17%)
%nec 62% 13% 17%
PD 5 (6%) 4 (4%) 4 (4%)
%nec 25% 7% 7%
pathologist by comparing pre-LT imaging and macro-
scopic analysis of the explanted liver. All macroscopi-
cally identified nodules were submitted for microscopic
analysis. The entire nodule was submitted, and the
presence of necrosis was visually semiquantitatively
classified as major necrosis in the presence of >90% of
necrosis because studies have reported increased overall
survival and recurrence-free survival after LT above
this cutoff.(17-20) Moreover, in a recent series of
patients with HCC who underwent TACE prior to
liver resection or transplantation, a pathological
response >90% has also been shown to be a significant
and independent factor associated with better overall
survival and recurrence-free survival.(5) Necrosis
between 50% and 90% and <50% were defined as
intermediate and minor necrosis, respectively. The
presence of microvascular invasion and satellite nodules
were recorded for each nodule. In case of residual
tumors, the grade of differentiation was assessed.
Statistical Analysis
Data were expressed as descriptive statistics (mean 6
standard deviation and range, or n [%], as appropriate)
and compared by the Pearson’s chi-square for categori-
cal data and 1-way analysis of variance for continuous
data. The influence of the different clinical variables,
tumoral differences, and features of the TACE proce-
dure on the presence of major tumor necrosis was ana-
lyzed. Variables associated with major tumor necrosis
with P < 0.05 in univariate analysis were entered into a
stepwise logistic regression model (conditional
Results
PATIENT AND TUMOR
CHARACTERISTICS AT BASELINE
Mean age at LT was 54 years (range, 31-64 years).
The causes of cirrhosis were alcohol-related (n 5 22),
hepatitis C virus (HCV; n 5 19), hepatitis B virus
(n 5 13), and other (n 5 4). The mean physiological
Model for End-Stage Liver Disease score was 12
(range 7-27). Before TACE, patients had a mean of
1.6 (range, 1-4). Tumors were a mean 25 mm in diam-
eter (range, 10-80 mm). HCC was solitary in 30 (52%)
patients. Seventy-one (80%) lesions presented with
washout and 25 (28%) with a peripheral capsule. Two
(2%) HCCs were associated with bland portal vein
thrombosis. There were no imaging features sugges-
ting tumoral portal vein obstruction.
TACE PROCEDURES
Patients underwent 94 TACE procedures, with a
mean 1.6 sessions per patient (range, 1-3). Twenty-
eight (48%) patients underwent 1 treatment session,
whereas 24 (41%) and 6 (11%) patients underwent 2
and 3 sessions, respectively. Retreatment was indicated
when viable tumor tissue (ie, tumor enhancement on
arterial-phase CT) was seen after the first session of
TACE. Arterial catheterization was selective in 64
(68%) procedures and nonselective in the remaining 30
(32%).
PATHOLOGICAL ANALYSIS
Among 88 nodules, 51 (58%), 7 (8%), and 30 (34%)
showed major (>90%), intermediate (50%-90%), and
minor (<50%) necrosis on pathology, respectively.
Forty-one (46%) nodules showed complete (100%)
necrosis. Microvascular invasion and the presence of

ORIGINAL ARTICLE | 1495

 15276473, 2016, 11, Downloaded from https://aasldpubs.onlinelibrary.wiley.com/doi/10.1002/lt.24615 by Universita Di Firenze Sistema, Wiley Online Library on [24/07/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
BURGIO ET AL. LIVER TRANSPLANTATION, November 2016

TABLE 2. Correlation Between Tumor Response According to mRECIST and EASL Criteria and the Rate of Tumor
Necrosis on Pathology
Reader 1 Reader 2
Major Intermediate Minor Major Intermediate Minor
Necrosis Necrosis Necrosis Necrosis Necrosis Necrosis
(n 5 51) (n 5 7) (n 5 30) (n 5 51) (n 5 7) (n 5 30)
mRECIST
CR
Overall 47 (92) 4 (57) 8 (27) 45 (88) 5 (71) 9 (30)
Complete lipiodol deposition 32 (63) 2 (29) 2 (7) 32 (63) 2 (29) 2 (7)
Incomplete lipiodol deposition 15 (29) 2 (29) 6 (20) 13 (25) 3 (43) 7 (23)
PR 3 (6) 2 (29) 7 (23) 6 (12) 1 (14) 2 (7)
SD 1 (2) 1 (14) 11 (37) 0 (0) 1 (14) 11 (37)
PD 0 (0) 0 (0) 4 (13) 0 (0) 0 (0) 4 (13)
EASL
CR
Overall 47 (92) 4 (57) 8 (27) 45 (88) 5 (71) 9 (30)
Complete lipiodol deposition 32 (63) 2 (29) 2 (7) 32 (63) 2 (29) 2 (7)
Incomplete lipiodol deposition 15 (29) 2 (29) 6 (20) 13 (25) 3 (43) 7 (23)
PR 2 (4) 2 (29) 5 (17) 5 (10) 1 (14) 4 (13)
SD 2 (4) 1 (14) 13 (43) 1 (2) 1 (14) 13 (43)
PD 0 (0) 0 (0) 4 (13) 0 (0) 0 (0) 4 (13)

NOTE: In nodules classified as CR, those with complete and incomplete lipiodol deposition are separated. Data are given as n (%).

satellite nodules were present in 25 (29%) and 21
(24%) HCCs, respectively. In residually viable HCCs,
tumor differentiation was classified as well, moderately,
and poorly differentiated in 17 (36%), 25 (53%), and 5
(11%), respectively.
CORRELATION BETWEEN
TUMOR NECROSIS AND
RESPONSE CRITERIA
Tumor response according to the different criteria, as
well as the mean tumor necrosis on pathology for each
criterion are summarized in Tables 1 and 2. Tumor
response according to the tumor size is provided in
Supporting Table 1.
The mean rate of tumor necrosis on pathology
decreased from complete response (CR) to progressive
disease (PD) for each criterion.
When considering the 51 nodules with major
necrosis upon pathologic analysis, 2 (4%), 47 (92%),
and 47 (92%) were classified as CR according to
RECIST 1.1, mRECIST, and EASL, respectively, for
reader 1; 1 (2%), 45 (88%), and 45 (88%) were classi-
fied as CR according to RECIST 1.1, mRECIST, and
EASL, respectively, for reader 2.
Among the 59 nodules classified as a CR according
to mRECIST or EASL for readers 1 and 2, 47 (80%)
and 45 (76%) had a major necrosis on pathological
examination, respectively.
Table 3 shows the tumor response according to
RECIST 1.1, mRECIST, and EASL criteria for
lesions with either satellite nodules or microvascular
invasion. Supporting Table 2 shows the correlation
between tumor response according to mRECIST and
EASL criteria and tumor differentiation on pathology
in nodules with remnant viable tumor cells after
TACE.
We have also compared nodules with 90%-99% and
those with 100% necrosis upon pathologic analysis.
There was no significant difference concerning the rate
of incomplete lipiodol deposition (2/10 [20%] versus
8/41 [20%]; P 5 1.00), the presence of satellite nodules
(1/10 [10%] versus 8/41 [20%]; P 5 0.66), and the rate
of microvascular invasion (2/10 [20%] versus 4/41
[10%]; P 5 0.58).
We only found that nodules with 100% necrosis
were smaller than those with 90%-99% necrosis
(mean, 35 6 18 mm versus 23 6 13 mm; P 5 0.02).
INFLUENCE OF LIPIODOL
DEPOSITION
Among the 59 nodules classified with a CR according
to mRECIST or EASL, those presenting with com-
plete lipiodol deposition (n 5 36, 61%) had a signifi-
cantly higher rate of necrosis than those with
incomplete lipiodol deposition (n 5 23, 39%): 95%
versus 68% (P 5 0.002) and 95% versus 63%
(P < 0.001) for readers 1 and 2, respectively (Fig. 2).

1496 | ORIGINAL ARTICLE

 15276473, 2016, 11, Downloaded from https://aasldpubs.onlinelibrary.wiley.com/doi/10.1002/lt.24615 by Universita Di Firenze Sistema, Wiley Online Library on [24/07/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
LIVER TRANSPLANTATION, Vol. 22, No. 11, 2016 BURGIO ET AL.

TABLE 3. Correlation Between the Presence of Satellite Nodules and Microvascular Invasion With Tumor Assessment for
the 2 Readers and Pathological Assessment of Tumor Necrosis
Satellite Nodules (n 5 25) Microvascular Invasion (n 5 21)
CR PR SD PD CR PR SD PD
Reader 1
RECIST 0 (0%) 6 (24%) 15 (60%) 4 (16%) 0 (0%) 4 (19%) 13 (62%) 4 (19%)
mRECIST 14* (56%) 4 (19%) 5 (20%) 2 (8%) 9† (43%) 4 (19%) 5 (24%) 3 (14%)
EASL 14* (56%) 4 (16%) 5 (20%) 2 (8%) 9† (43%) 4 (19%) 5 (24%) 3 (14%)
Reader 2
RECIST 0 (0%) 9 (36%) 14 (56%) 2 (8%) 0 (0%) 5 (24%) 13 (62%) 3 (14%)
mRECIST 14* (56%) 4 (16%) 5 (20%) 2 (8%) 10‡ (48%) 4 (19%) 4 (19%) 3 (14%)
EASL 14* (56%) 4 (16%) 5 (20%) 2 (8%) 10‡ (48%) 4 (19%) 4 (19%) 3 (14%)
Pathologic examination
Major necrosis (> 90%) 9§ (36%) 6k (28%)
Intermediate necrosis (50%-90%) 5 (20%) 3 (14%)
Minor necrosis (< 50%) 11 (44%) 12 (57%)

*Including 4 nodules with complete lipiodol deposition.

†
Including 4 nodules with complete lipiodol deposition.
‡
Including 5 nodules with complete lipiodol deposition.
§
8 nodules showed 100% necrosis.
k
4 nodules showed 100% necrosis.

Among the 59 nodules classified as a CR, 35%
(8/23) and 43% (10/23) of those with incomplete lip-
iodol deposition had intermediate or minor necrosis
for readers 1 and 2, respectively. It was the case in only
11% (4/36) of those with complete lipiodol deposition
(P 5 0.04 and P 5 0.01 for reader 1 and 2, respectively;
Table 2).
Finally, CR nodules with or without complete lipio-
dol deposition had similar rates of satellite nodules
(25% versus 22%; P 5 1.00, for reader 1; 22% versus
26%; P 5 1.00, for reader 2), and microvascular inva-
sion (11% versus 22%; P 5 0.29, for reader 1; 11% ver-
sus 26%; P 5 0.16, for reader 2).
INTEROBSERVER AGREEMENT
FOR RESPONSE CRITERIA
The interobserver agreement for RECIST 1.1,
mRECIST, and EASL criteria was substantial
(j 5 0.65 6 0.08; j 5 0.78 6 0.07; and j 5 0.75 6
0.07), respectively.
FACTORS ASSOCIATED WITH
MAJOR NECROSIS
Besides tumor response to mRECIST or EASL crite-
ria (P < 0.001 for both readers), complete lipiodol
deposition (P < 0.001) was associated with major
tumor necrosis on univariate analysis (Table 4).
Only the response to mRECIST was statistically
associated with major necrosis for both readers by
logistic regression analysis (P < 0.001; odds ratio
[OR], 24; 95% confidence interval [CI], 7-82, for
reader 1; OR, 12.3; 95% CI, 4.2-36.3, for reader 2;
Table 4).
Discussion
This study compared tumor response following TACE
based on contrast-enhanced CT according to 3 differ-
ent response criteria (RECIST 1.1, mRECIST, and
EASL), to the percentage of necrosis on histological
analysis as the method of reference. We showed that
both mRECIST and EASL criteria are well correlated
with tumor necrosis compared with RECIST 1.1 and
that mRECIST slightly outperformed the EASL cri-
teria. We also showed that both criteria were
reproducible.
Selection criteria for LT in patients with HCC are
based on the Milan criteria, ie, tumor number and size.
The outcome in patients is excellent with an expected
70% 5-year OS rate.(21) Local bridging therapy while
on the waiting list for LT is performed in most coun-
tries because of the long wait on the list. Treatment is
mostly indicated to avoid patient dropout and to con-
trol tumor progression. Several studies have reported
the usefulness and efficacy of TACE as a bridging
therapy for patients on the LT waiting list.(22-25)
Another benefit of TACE is to induce tumor necrosis,
which is now known to be associated with both
improved survival and recurrence-free survival after
LT.(5,17-20) Thus, if the radiological assessment of
tumor response following TACE is correlated with the

ORIGINAL ARTICLE | 1497

 15276473, 2016, 11, Downloaded from https://aasldpubs.onlinelibrary.wiley.com/doi/10.1002/lt.24615 by Universita Di Firenze Sistema, Wiley Online Library on [24/07/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
BURGIO ET AL. LIVER TRANSPLANTATION, November 2016

TABLE 4. Tumor Features, TACE Procedure Characteristics, and Tumor Evaluation According to the Presence of Major
Tumor Necrosis (>90%) or Not
Major Necrosis P Value
Logistic
Yes (n 5 51) No (n 5 37) Univariate Regression*
Tumor features at baseline
Number of nodules
1 15 (29) 18 (49) 0.74
2 22 (43) 8 (22)
3 13 (25) 8 (21)
4 1 (2) 3 (8)
Size of nodules, mm 25 6 2.2 26 6 2.1 0.87
Bland portal vein thrombosis 1 (2) 1 (3) 0.66
Capsule 14 (27) 11 (30) 0.68
TACE procedure characteristics: 0.77
Number of TACE sessions per lesion
1 30 (59) 22 (59)
2 18 (35) 14 (38)
3 3 (6) 1 (3)
Selective treatment 35 (69) 25 (68) 0.54
Tumor evaluation on CT
Complete lipiodol deposition 32 (63) 4 (11) < 0.001
Response for reader 1
RECIST 1.1 0.01
CR 2 (4) 0 (0)
PR 19 (37) 5 (13)
SD 28 (55) 25 (68)
PD 2 (4) 7 (19)
mRECIST < 0.001 < 0.001
CR 47 (92) 12 (32)
PR 3 (6) 9 (24)
SD 1 (2) 12 (32)
PD 0 (0) 4 (11)
EASL < 0.001
CR 47 (92) 12 (32)
PR 2 (4) 7 (19)
SD 2 (4) 14 (38)
PD 0 (0) 4 (11)
Response for reader 2
RECIST 1.1 0.24
CR 1 (2) 0 (0)
PR 21 (41) 12 (32)
SD 28 (54) 21 (57)
PD 1 (2) 4 (11)
mRECIST
CR 45 (88) 14 (37)
PR 6 (12) 7 (19) < 0.001 < 0.001
SD 0 (0) 12 (32)
PD 0 (0) 4 (11)
EASL < 0.001
CR 45 (88) 14 (37)
PR 5 (10) 5 (13)
SD 1 (2) 14 (37)
PD 0 (0) 4 (11)

NOTE: mRECIST and EASL are entered into the logistic regression model as categorical variables. Data are given as n (%) or
mean 6 standard deviation.
*Multivariate logistic regression was performed twice with only 1 reader per analysis.

pathological response, this could be crucial in predict- independent factor of better overall and recurrence-
ing patient outcome, in selecting patients for LT, and free survival.
in adapting the delay to LT. Allard et al.(5) recently RECIST 1.1 criteria have been shown to markedly
showed that a pathologic response >90% is an underestimate tumor necrosis following locoregional

1498 | ORIGINAL ARTICLE


LIVER TRANSPLANTATION, Vol. 22, No. 11, 2016
therapies.(9) Indeed these criteria are only based on tumor
shrinkage after treatment. Our results support these data
as readers 1 and 2 classified only 2-4% of 51 nodules with
major necrosis on pathological analysis as CR according
to RECIST 1.1.
Unlike RECIST 1.1, mRECIST and EASL criteria
are based on an analysis of the viable portion of
tumors, defined as the areas of enhancement on
contrast-enhanced imaging. In our series, mRECIST
and EASL classified 88%-92% of the nodules with
major necrosis as having a CR, confirming the good
correlation of these criteria with the amount of tumor
necrosis.(26) It is important to note that an objective
response according to either mRECIST or EASL cri-
teria has been shown to be an independent prognostic
factor and associated with improved survival in patients
treated with TACE.(12-15) The good correlation
between the pathological response and mRECIST or
EASL criteria suggests that priority for LT should
take into account tumor response on imaging.
The main concern regarding the radiological evalua-
tion of tumor response following TACE is the possi-
bility of overestimating tumor necrosis as shown by
Bargellini et al.(16) In their series, pathological exami-
nation necrosis was found to be incomplete in 59/157
(37.6%) of nodules classified as having a CR on
contrast-enhanced CT. On the other hand, underesti-
mation was found to be much more uncommon
because only 3/72 (4.2%) nodules classified as stable
disease (SD) were found to have necrosis >50%. This
was confirmed by the present results. Indeed, among
the 59 nodules classified with a CR by mRECIST or
EASL criteria by the 2 readers, 12-14 had intermedi-
ate or minor necrosis (ie, < 90%) leading to an overes-
timation in 20%-24% of the patients, respectively.
Conversely, underestimation of necrosis by either
mRECIST or EASL criteria was not frequent as only
8%-12% of nodules with major necrosis were classified
as partial response (PR) or SD by the 2 readers.
These errors are often explained by the presence of lip-
iodol deposits, which make depiction of viable enhanced
tissue difficult.(27,28) They could probably be explained by
an incorrect analysis of the lipiodol deposition. It is inter-
esting to note that our series showed that the presence of
complete lipiodol deposition was associated with the
highest rate of tumor necrosis (mean 95% for both read-
ers), which supports previously published data.(29,30)
Moreover, when lack of enhancement (nodules classified
as CR by either mRECIST or EASL) was combined
with complete lipiodol deposition, overestimation of
tumor necrosis was reduced to 11%, suggesting that these
15276473, 2016, 11, Downloaded from https://aasldpubs.onlinelibrary.wiley.com/doi/10.1002/lt.24615 by Universita Di Firenze Sistema, Wiley Online Library on [24/07/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
BURGIO ET AL.
2 imaging findings should be combined to improve eval-
uation of the pathological response. On the other hand,
this also suggests that care should be taken when classify-
ing a nodule with incomplete lipiodol deposition as CR
by mRECIST or EASL because 35%-43% of them have
an intermediate or minor pathological response.
Results of mRECIST and EASL criteria were simi-
lar. All nodules classified as CR by 1 criterion were
also classified as CR by the other. This was not surpris-
ing because our population only included patients who
were undergoing LT, and who therefore had a few
small tumors limiting potential disagreement. Similar
results were reported in another series.(31) Yet, on mul-
tivariate analysis, mRECIST slightly outperformed
EASL for both readers, probably due to a better alloca-
tion of patients with intermediate necrosis.
This study has certain limitations besides the retro-
spective design. First, evaluation of the tumor response
was evaluated on the last CT before LT. Thus, the
delay between the TACE procedures and the assess-
ment of tumor response varied among patients, but
remained fairly short (no longer than 3 months). Sec-
ond, survival data were not evaluated, but the aim of
this study was to evaluate and compare the ability of
different imaging criteria to predict the extent of
tumoral necrosis. Third, magnetic resonance imaging
might be more sensitive than CT in depicting incom-
plete treatment/focal recurrence because depiction of
residual tumor is not hampered by lipiodol deposition,
but this has not been demonstrated in large studies yet.
Because most teams use routine CT to assess tumor
response after conventional TACE, we think it is clini-
cally relevant to evaluate CT in those patients.
In conclusion, both mRECIST and EASL criteria
are reproducible and have a better correlation with
pathological necrosis than RECIST 1.1 criteria. The
combination of CR with the mRECIST or EASL cri-
teria and complete lipiodol deposition is associated
with the highest rate of necrosis and reduces overesti-
mation of tumor necrosis. Even in nodules considered
to be CR according to mRECIST or EASL criteria,
an incomplete lipiodol deposition should be considered
indicative of substantial viable tumor remnant.
REFERENCES
1) Jemal A, Bray F, Center MM, Ferlay J, Ward E, Forman D.
Global cancer statistics. CA Cancer J Clin 2011;61:69-90.
2) Bruix J, Sherman M; for American Association for the Study of
Liver Diseases. Management of hepatocellular carcinoma: an
update. Hepatology 2011;53:1020-1022.
ORIGINAL ARTICLE | 1499

BURGIO ET AL.
3) European Association For The Study Of The Liver; European
Organisation For Research And Treatment Of Cancer. EASL-
EORTC clinical practice guidelines: management of hepatocellu-
lar carcinoma. J Hepatol 2012;56:908-943.
4) Dharancy S, Boitard J, Decaens T, Sergent G, Boleslawski E,
Duvoux C, et al. Comparison of two techniques of transarterial
chemoembolization before liver transplantation for hepatocellular
carcinoma: a case-control study. Liver Transpl 2007;13:665-671.
5) Allard MA, Sebagh M, Ruiz A, Guettier C, Paule B, Vibert E,
et al. Does pathological response after transarterial chemoemboli-
zation for hepatocellular carcinoma in cirrhotic patients with cir-
rhosis predict outcome after liver resection or transplantation?
J Hepatol 2015;63:83-92.
6) Cescon M, Cucchetti A, Ravaioli M, Pinna AD. Hepatocellular
carcinoma locoregional therapies for patients in the waiting list.
Impact on transplantability and recurrence rate. J Hepatol 2013;
58:609-618.
7) Otto G, Herber S, Heise M, Lohse AW, M€onch C, Bittinger F,
et al. Response to transarterial chemoembolization as a biological
selection criterion for liver transplantation in hepatocellular carci-
noma. Liver Transpl 2006;12:1260-1267.
8) Eisenhauer EA, Therasse P, Bogaerts J, Schwartz LH, Sargent
D, Ford R, et al. New response evaluation criteria in solid
tumours: revised RECIST guideline (version 1.1). Eur J Cancer
2009;45:228-247.
9) Forner A, Ayuso C, Varela M, Rimola J, Hessheimer AJ, de
Lope CR, et al. Evaluation of tumor response after locoregional
therapies in hepatocellular carcinoma: are response evaluation cri-
teria in solid tumors reliable? Cancer 2009;115:616-623.
10) Bruix J, Sherman M, Llovet JM, Beaugrand M, Lencioni R,
Burroughs AK, et al.; for EASL Panel of Experts on HCC.
Clinical management of hepatocellular carcinoma. Conclusions of
the Barcelona-2000 EASL conference. European Association for
the Study of the Liver. J Hepatol 2001;35:421-430.
11) Lencioni R, Llovet JM. Modified RECIST (mRECIST) assess-
ment for hepatocellular carcinoma. Semin Liver Dis 2010;30:52-60.
12) Gillmore R, Stuart S, Kirkwood A, Hameeduddin A, Woodward
N, Burroughs AK, Meyer T. EASL and mRECIST responses
are independent prognostic factors for survival in hepatocellular
cancer patients treated with transarterial embolization. J Hepatol
2011;55:1309-1316.
13) Bargellini I, Vignali C, Cioni R, Petruzzi P, Cicorelli A,
Campani D, et al. Hepatocellular carcinoma: CT for tumor
response after transarterial chemoembolization in patients exceed-
ing Milan criteria--selection parameter for liver transplantation.
Radiology 2010;255:289-300.
14) Jung ES, Kim JH, Yoon EL, Lee HJ, Lee SJ, Suh SJ, et al.
Comparison of the methods for tumor response assessment in
patients with hepatocellular carcinoma undergoing transarterial
chemoembolization. J Hepatol 2013;58:1181-1187.
15) Memon K, Kulik L, Lewandowski RJ, Wang E, Riaz A, Ryu
RK, et al. Radiographic response to locoregional therapy in hepa-
tocellular carcinoma predicts patient survival times. Gastroenter-
ology 2011;141:526-535.
16) Bargellini I, Bozzi E, Campani D, Carrai P, De Simone P,
Pollina L, et al. Modified RECIST to assess tumor response
after transarterial chemoembolization of hepatocellular carcinoma:
CT-pathologic correlation in 178 liver explants. Eur J Radiol
2013;82:e212-e218.
17) Ho MH, Yu CY, Chung KP, Chen TW, Chu HC, Lin CK,
Hsieh CB. Locoregional therapy-induced tumor necrosis as a
1500 | ORIGINAL ARTICLE
15276473, 2016, 11, Downloaded from https://aasldpubs.onlinelibrary.wiley.com/doi/10.1002/lt.24615 by Universita Di Firenze Sistema, Wiley Online Library on [24/07/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
LIVER TRANSPLANTATION, November 2016
predictor of recurrence after liver transplant in patients with
hepatocellular carcinoma. Ann Surg Oncol 2011;18:3632-3639.
18) Chan KM, Yu MC, Chou HS, Wu TJ, Lee CF, Lee WC. Sig-
nificance of tumor necrosis for outcome of patients with hepato-
cellular carcinoma receiving locoregional therapy prior to liver
transplantation. Ann Surg Oncol 2011;18:2638-2646.
19) Kornberg A, Witt U, K€ upper B, Wildgruber M, Friess H. Post-
interventional tumor necrosis predicts recurrence-free long-term
survival in liver transplant patients with advanced hepatocellular
carcinoma. Transplant Proc 2013;45:1913-1915.
20) Cantu M, Piardi T, Sommacale D, Ellero B, Woehl-Jaegle ML,
Audet M, et al. Pathologic response to non-surgical locoregional
therapies as potential selection criteria for liver transplantation for
hepatocellular carcinoma. Ann Transplant 2013;18:273-284.
21) Mazzaferro V, Bhoori S, Sposito C, Bongini M, Langer M,
Miceli R, Mariani L. Milan criteria in liver transplantation for
hepatocellular carcinoma: an evidence-based analysis of 15 years
of experience. Liver Transpl 2011;17(suppl 2):S44-S57.
22) Seehofer D, Nebrig M, Denecke T, Kroencke T, Weichert W,
Stockmann M, et al. Impact of neoadjuvant transarterial chemo-
embolization on tumor recurrence and patient survival after liver
transplantation for hepatocellular carcinoma: a retrospective anal-
ysis. Clin Transplant 2012;26:764-774.
23) Frangakis C, Geschwind JF, Kim D, Chen Y, Koteish A, Hong
K, et al. Chemoembolization decreases drop-off risk of hepato-
cellular carcinoma patients on the liver transplant list. Cardiovasc
Intervent Radiol 2011;34:1254-1261.
24) Chapman WC, Majella Doyle MB, Stuart JE, Vachharajani N,
Crippin JS, Anderson CD, et al. Outcomes of neoadjuvant trans-
arterial chemoembolization to downstage hepatocellular carcino-
ma before liver transplantation. Ann Surg 2008;248:617-625.
25) Graziadei IW, Sandmueller H, Waldenberger P, Koenigsrainer
A, Nachbaur K, Jaschke W, et al. Chemoembolization followed
by liver transplantation for hepatocellular carcinoma impedes
tumor progression while on the waiting list and leads to excellent
outcome. Liver Transpl 2003;9:557-563.
26) Riaz A, Memon K, Miller FH, Nikolaidis P, Kulik LM,
Lewandowski RJ, et al. Role of the EASL, RECIST, and
WHO response guidelines alone or in combination for hepato-
cellular carcinoma: radiologic-pathologic correlation. J Hepatol
2011;54:695-704.
27) Kubota K, Hisa N, Nishikawa T, Fujiwara Y, Murata Y, Itoh S,
et al. Evaluation of hepatocellular carcinoma after treatment with
transcatheter arterial chemoembolization: comparison of
Lipiodol-CT, power Doppler sonography, and dynamic MRI.
Abdom Imaging 2001;26:184-190.
28) Yamashita Y, Yoshimatsu S, Sumi M, Harada M, Takahashi M.
Dynamic MR imaging of hepatoma treated by transcatheter arte-
rial embolization therapy. Assessment of treatment effect. Acta
Radiol 1993;34:303-308.
29) Kwan SW, Fidelman N, Ma E, Kerlan RK Jr, Yao FY. Imaging
predictors of the response to transarterial chemoembolization in
patients with hepatocellular carcinoma: a radiological-pathological
correlation. Liver Transpl 2012;18:727-736.
30) Shim JH, Han S, Shin YM, Yu E, Park W, Kim KM, et al. Optimal
measurement modality and method for evaluation of responses to
transarterial chemoembolization of hepatocellular carcinoma based on
enhancement criteria. J Vasc Interv Radiol 2013;24:316-325.
31) Shim JH, Lee HC, Kim SO, Shin YM, Kim KM, Lim YS, Suh
DJ. Which response criteria best help predict survival of patients
with hepatocellular carcinoma following chemoembolization. A vali-
dation study of old and new models. Radiology 2012;262:708-718.