Transcript Gastrointestinal Cancer Committee

2 or 3 Year DFS is an
Appropriate Primary Endpoint in
Stage III Adjuvant Colon Cancer
Trials with Fluoropyrimidines
with or without Oxaliplatin or
Irinotecan
D Sargent, G Yothers, E Van Cutsem, J
Cassidy, L Saltz, N Wolmark, B Bot, Q
Shi, M Buyse, and A de Gramont for the
Adjuvant Colon Cancer Endpoints
(ACCENT) Group
ASCO 2009
Disclosure
Consulting/Honoraria
• Pfizer
• Sanofi
• Genentech
• Amgen
• Roche
for work unrelated to the current project
The Adjuvant Colon Cancer
Endpoints (ACCENT) database
• Established in 2003, to validate diseasefree survival (DFS) as an endpoint in
adjuvant colon cancer
• Originally included individual patient
data from 18 large adjuvant clinical
trials, 21,000 pts
• Jointly owned by all contributors
Previous ACCENT findings
• DFS surrogate for OS1
• Concordance higher in stage III vs II2
• 2 yr DFS may be adequate2
• Post-recurrent OS ↑ for stage II vs III3
• Majority of rx benefit in first 2 years4
• At least 6 yrs for OS benefit in new trials5
et al, JCO 2005; 2Sargent et al JCO 2007; 3O’Connell et al JCO 2008
4Sargent et al JCO 2009; 5DeGramont et al ASCO 2008
1Sargent
ACCENT update: 6 adjuvant
trials added
Trial
MOSAIC
X-ACT
NSABP C-06
NSABP C-07
CALGB 89803
PETACC-3
Accrual
Period
1998-01
1998-01
1997-99
2000-02
1999-01
2000-02
# patients
Experimental
treatment arm
% stage
III
2246
1987
1557
2434
1264
3188
FOLFOX4
Capecitabine
Uracil/tegafur
FLOX
IFL
FOLFIRI
60
100
53
71
100
71
• Total addition 12,676 new patients
• Median follow-up on living patients: 6 years
• Median survival following recurrence: 20 months
Goals of new analysis
• Confirm association between DFS & OS in
new trials
• Validate that the DFS/OS association is
stronger in pts with stage III than II disease
• Validate 2 yr DFS time-point
• Explore duration of OS follow-up required to
demonstrate surrogacy (5 v 6 or more yrs)
Methods
• Included only patients from new trials not
included in previous ACCENT analyses
• Graphical representation of hazard functions
• Association of within trial hazard ratios
•
•
•
•
Compare DFS, OS btwn arms within each study
Landmarks: 2, 3 yr for DFS, 5, 6 yr for OS
Weighted regression (WLS) – simple
Bivariate survival analysis1 (Copula) - complex
• Goal: All methods give consistent results
1Burzykowski
JRSS-C 2001
Recurrence rate over time
9
7.64
8
Recurrence Rate (%)
7
83% of recurrences
occur within the
first 3 years
6.92
6
5.44
5
4
3.68
3
2.97
2.63
2.07
2
1.7
1.32 1.23
0.86
1
0.6
0.14
0
0
1
2
3
Years
4
5
6
DFS: experimental vs control
pooled across trials
0.0006
Hazard Rate
0.0004
Experimental
Control
0.0002
0
0
2
4
Follow-up Time (Years)
6
Within trial hazard ratios for
3 year DFS vs 5 year OS
3 Year DFS v. 5 Year OS
1.2
OS Hazard Ratio
1.1
1
0.9
WLS R2 = 0.60
0.8
Copula R2 = 0.20
0.7
0.7
0.8
0.9
1
DFS Hazard Ratio
1.1
1.2
Within trial hazard ratios for
3 year DFS vs 5 & 6 year OS
3 Year DFS v. 6 Year OS
1.2
1.2
1.1
1.1
OS Hazard Ratio
OS Hazard Ratio
3 Year DFS v. 5 Year OS
1
0.9
WLS R2 = 0.60
0.8
1
0.9
WLS R2 = 0.75
0.8
Copula R2 = 0.20
Copula R2 = 0.17
0.7
0.7
0.7
0.8
0.9
1
DFS Hazard Ratio
1.1
1.2
0.7
0.8
0.9
1
DFS Hazard Ratio
1.1
1.2
Within trial hazard ratios for
2 year DFS vs 5 year OS
2 Year DFS v. 5 Year OS
1.2
OS Hazard Ratio
1.1
1
0.9
WLS R2 = 0.58
0.8
Copula R2 = 0.37
0.7
0.7
0.8
0.9
1
DFS Hazard Ratio
1.1
1.2
Within trial hazard ratios for
2 year DFS vs 5 & 6 year OS
2 Year DFS v. 6 Year OS
1.2
1.2
1.1
1.1
OS Hazard Ratio
OS Hazard Ratio
2 Year DFS v. 5 Year OS
1
0.9
WLS R2 = 0.58
0.8
1
0.9
WLS R2 = 0.76
0.8
Copula R2 = 0.37
Copula R2 = 0.49
0.7
0.7
0.7
0.8
0.9
1
DFS Hazard Ratio
1.1
1.2
0.7
0.8
0.9
1
DFS Hazard Ratio
1.1
1.2
Conclusions: DFS as an
endpoint in joint stage II/III trials
DFS yrs
2
2
OS yrs
5
6
WLS R2
0.58
0.76
Copula R2
0.37
0.49
3
3
5
6
0.60
0.75
0.20
0.17
• Association higher for DFS with 6 vs 5 yr OS
• Associations with 2 and 3 yr DFS and OS similar
• Overall, modest associations
Stage III within trial HR
3 year DFS v 5 year OS
Stage III only - 3 Year DFS v. 5
Year OS
OS Hazard Ratio
1.2
1.1
1
0.9
WLS R2 = 0.93
0.8
Copula R2 = 0.81
0.7
0.7
0.8
0.9
1
DFS Hazard Ratio
1.1
1.2
Stage III within trial HR
3 year DFS v 5 & 6 year OS
Stage III only - 3 Year DFS v. 6
Year OS
1.2
1.2
1.1
1.1
OS Hazard Ratio
OS Hazard Ratio
Stage III only - 3 Year DFS v. 5
Year OS
1
0.9
WLS R2 = 0.93
0.8
1
0.9
WLS R2 = 0.87
0.8
Copula R2 = 0.81
Copula R2 = 0.79
0.7
0.7
0.7
0.8
0.9
1
DFS Hazard Ratio
1.1
1.2
0.7
0.8
0.9
1
DFS Hazard Ratio
1.1
1.2
Stage III within trial HR
2 year DFS v 5 year OS
Stage III only - 2 Year DFS v. 5
Year OS
OS Hazard Ratio
1.2
1.1
1
0.9
WLS R2 = 0.91
0.8
Copula R2 = 0.86
0.7
0.7
0.8
0.9
1
DFS Hazard Ratio
1.1
1.2
Stage III within trial HR
2 year DFS v 5 & 6 year OS
Stage III only - 2 Year DFS v. 6
Year OS
1.2
1.2
1.1
1.1
OS Hazard Ratio
OS Hazard Ratio
Stage III only - 2 Year DFS v. 5
Year OS
1
0.9
WLS R2 = 0.91
0.8
1
0.9
WLS R2 = 0.93
0.8
Copula R2 = 0.86
Copula R2 = 0.88
0.7
0.7
0.7
0.8
0.9
1
DFS Hazard Ratio
1.1
1.2
0.7
0.8
0.9
1
DFS Hazard Ratio
1.1
1.2
Conclusions: DFS as a stage III
endpoint
DFS yrs
2
2
OS yrs
5
6
WLS R2
0.91
0.93
Copula R2
0.86
0.88
3
3
5
6
0.93
0.87
0.81
0.79
• Concordance high for DFS with both 5 & 6 yr OS
• DFS at 2 and 3 years very similar
• Strong association
Forest plot: 2 Yr DFS v 6 Yr OS
stage III
Trial #1
DFS
OS
Trial #2
Trial #3
Trial #4
Trial #5
Trial #6
Overall
0
0.2
0.4
0.6
0.8
1
1.2
1.4
1.6
How well did previous ACCENT
model predict these results?
• Prior analyses testing only 5-FU-based
treatment produced a predictive model to
use early DFS to predict later OS
• Used observed early DFS HRs from the
new trials, and the previous model, to
predict OS HRs in the 6 new trials
Prediction plot: 2 Yr DFS v 6 Yr
OS: stage III
1.3
Predicted HR
HR: 6 year OS
1.2
1.1
1
0.9
0.8
0.7
0.7
0.8
0.9
1
HR: 2 year DFS
1.1
1.2
Prediction plot: 2 Yr DFS v 6 Yr
OS: stage III
1.3
Predicted HR
Actual HR
HR: 6 year OS
1.2
1.1
1
0.9
0.8
0.7
0.7
0.8
0.9
1
HR: 2 year DFS
1.1
1.2
Discussion
• Follow-up beyond 6 years not available
at this time
• All trials used cytotoxic drugs, no
biologics
• If new agents change pattern of cancer
recurrence (delay vs prevent),
relationship between DFS and OS could
change
Validation of previous ACCENT
findings?
• DFS surrogate for OS – Confirmed (with
conditions)
• Concordance higher in stage III vs II Confirmed
• 2 yr DFS may be adequate - Confirmed
• > 6 yrs for OS benefit in new joint stage
II/III trials – Requires further confirmation
Conclusions
• Based on modern adjuvant colon
cancer trials
• In joint stage II & III trials, even 6 yrs
follow-up demonstrates only modest
association between DFS and OS
• In stage III patients, trial level DFS &
OS association remains strong
• Association between 2 yr DFS as
strong as 3 yr DFS with OS
Conclusion
DFS assessed after 2 or 3 years
remains an appropriate primary
endpoint for stage III adjuvant
colon cancer trials
ACCENT collaborators
S Wieand, G Yothers, M O’Connell, N Wolmark – NSABP
J Benedetti, C Blanke – SWOG
R Labianca – Ospedali Riuniti (Italy)
D Haller, P Catalano, A Benson – ECOG
C O’Callaghan – NCIC
JF Seitz – University of the Mediterranean (France)
G Francini – University of Siena (Italy)
A de Gramont, T Andre – GERCOR
R Goldberg, L Saltz, J Meyerhardt, N Jackson – CALGB
M Buyse – IDDI (Belgium)
R Gray, D Kerr – QUASAR
A Grothey, S Alberts, B Bot, E Green, Q Shi –Mayo Clinic
C Twelves -University of Bradford (UK)
J Cassidy – University of Glasgow (UK)
F Sirzen – Roche ; L Cisar - Pfizer
E Van Cutsem –University Hospital Gasthuisberg (Belgium);
A Sobrero - Ospedale San Martino (Italy)