BLINDED INDEPENDENT CENTRAL RESPONSE ASSESSMENT
USING RECIST, MODIFIED RECIST, AND CHOI CRITERIA
IN PATIENTS TREATED WITH SORAFENIB FOR ADVANCED
HEPATOCELLULAR CARCINOMA
Mohamed Bouattour, Johanna Wassermann, Onorina Bruno,
Béatrice Larroque, Laurent Castera, Chantal Dreyer, Valérie Vilgrain,
Jacques Belghiti, Eric Raymond, Sandrine Faivre
Beaujon University Hospital
Clichy, France
AASLD 2011
No disclosure to declare
AASLD 2011
RECIST do Not Capture the True Benefit of
Sorafenib in HCC
• Sorafenib improves survival but yields low objective
response rate by RECIST (< 5%)1,2
• Sustained survival despite NO response by RECIST
suggest that RECIST are inappropriate to capture the true
benefit of sorafenib3
• Modified RECIST (mRECIST)4 are used to assess vascular
effects of TACE5 and CHOI criteria were proposed to
evaluate necrosis induced by targeted agents in HCC6
1 Llovet JM ; N Engl J Med 2008 2 Cheng AL ; Lancet Oncol 2009 3 Edeline J, Cancer 2011
AASLD 2011
4 Lencioni R ; Semin Liver Dis. 2010 5 Gillmore R, J Hepatol 2011 6 Faivre S, Clin Cancer Res 2011
Objectives
• At the first tumor evaluation, are mRECIST and CHOI
criteria predicting overall survival in patients with
HCC treated with sorafenib?
• Can mRECIST and CHOI criteria reallocate to the
objective response group, patients who were
inappropriately considered non-responders by
RECIST?
AASLD 2011
Study Method/Radiological Evaluation
• Retrospective single center cohort analysis
• January 2007- December 2009
• Beaujon Hospital
• Baseline evaluation within 6 weeks prior to sorafenib
• First tumor evaluation by CT-scan 2-3 months after
sorafenib initiation
AASLD 2011
Radiological Evaluation
• Quality control criteria
– Multiphasic CT-scan fully available for central review
– Central Review of data by a radiologist highly experienced in
liver cancers, blinded to clinical data
– Evaluation of tumor response by RECIST, mRECIST, and CHOI
criteria
AASLD 2011
RECIST mRECIST CHOI criteria
HU
Results: Study Population Selection
Patients with BCLC B-C, Child-Pugh A-B, advanced
hepatocellular carcinoma treated with sorafenib
from 2007 to 2009
(n=82)
Non-evaluable patients (n=22) – Non-evaluable CT scan (n=9)
– Target lesions in pretreated area (n=9)
– No target lesion (n=3)
– Missing data (n=4) • Lost of follow up
• Early death
Patients evaluated in this study
(n=60)
AASLD 2011
Patient Characteristics
Median age, years 61 (37-77)
Sex M/F 52/8
Etiology, % (number of patients)
Viral
Alcohol
48 (29)
23 (14)
Child Pugh Score, % (number of patients)
A
B
80 (48)
20 (12)
BCLC stage , % (number of patients)
B
C
33 (20)
67 (40)
Pathological diagnosis, % (number of patients) 88 (53)
Extrahepatic Spread , % (number of patients) 35 (21)
Prior treatments, % (number of patients)
None
Surgery
Radio Frequency Ablation
Trans-Arterial Chemo-Embolization
32 (19)
27 (16)
5 (3)
36 (22)
Median duration of sorafenib, months 5.7
Median time for the first evaluation, months 2.1
AASLD 2011
Venn diagram of tumor response according to three criteria
2 10 15 RECIST n=2
mRECIST n=12
CHOI criteria n=27
mRECIST and CHOI Criteria Identify
More Responder Patients Than RECIST
AASLD 2011
AASLD 2011
0
10
20
30
40
50
60
70
RECIST mRECIST CHOI
3
21
45
62
48
23
3530
32
Response Rates by
RECIST, mRECIST, and CHOI Criteria
Progressive Disease
Stable Disease
Objective response
% o
f re
sp
on
se
rate
s b
y e
ac
h c
rite
ria
N=60 N=60 N=56* *4 pts non evaluable
At the first tumor evaluation, are mRECIST
and CHOI criteria predicting overall survival
in patients with HCC treated with sorafenib?
AASLD 2011
Responses by RECIST Criteria Correlate with
Survival
0
0,1
0,2
0,3
0,4
0,5
0,6
0,7
0,8
0,9
1
0 5 10 15 20 25 30 35 40 45
Objective response (PR/CR)
Stable Disease
Progressive Disease
RECIST
p=0.0012
Pro
bab
ilit
y o
f su
rviv
al
Duration of survival, months
Can mRECIST and CHOI criteria reallocate
to the objective response group, patients
who were inappropriately considered non-
responders by RECIST?
AASLD 2011
Examples of discrepancies
Between methods of evaluation
Baseline Stable
RECIST
Response
mRECIST
Response
CHOI criteria
↘ HU
Baseline Progression Response Response
↘ HU
AASLD 2011
0
5
10
15
20
25
30
0
5
10
15
20
25
30
0
5
10
15
20
25
30
Response by RECIST
Progressive Disease
Stable Disease
Objective response
Responders
RECIST (n=2)
Responders
mRECIST (n=12)
Responders
CHOI (n=27)
Many Stable and Some Progressive Diseases
by RECIST Are Objective Responses
by mRECIST and CHOI Criteria
Objective responses according to each criteria
Nu
mb
er
of
pa
tie
nts
wit
h o
bje
cti
ve
re
sp
on
se
2 2 2
9
1 22
3
3.3%
21.4%
45%
AASLD 2011
Conclusion
• Response rate using mRECIST and CHOI criteria correlates with survival in advanced HCC patients treated with sorafenib
• mRECIST and CHOI criteria identify patients with true benefit (partial responders with higher survival) among “RECIST-stable” and “RECIST-progressive” patients
• mRECIST and CHOI criteria compared to RECIST increase the number of partial responders who also are patients with a median overall survival >14 months
AASLD 2011