Randomized phase III study of capecitabine, oxaliplatin and
bevacizumab with or without cetuximab in advanced colorectal cancer
CAIRO2 study of the Dutch Colorectal Cancer Group (DCCG)
CJA Punt, J Tol, CJ Rodenburg, A Cats, GJ Creemers, JG Schrama, FLG Erdkamp, A Vos, L Mol, NF Antonini
Treatment of advanced colorectal cancer (CRC)
• Fluoropyrimidine-based chemotherapy plus the VEGF antibody bevacizumab is currently considered as the standard 1st line treatment
• Cetuximab is a chimaeric monoclonal antibody against the epidermal growth factor receptor (EGFR) which has shown efficacy as a single agent and in combination with chemotherapy
Background
• VEGF and EGFR share common downstream signaling pathways, and preclinical models have shown additive effects for VEGF and EGFR inhibition
• In irinotecan-resistant CRC the combination of irinotecan, cetuximab, and bevacizumab (BOND2 study)1 was feasible and suggested greater efficacy compared to irinotecan + cetuximab (BOND study)2
• Therefore, targeting both VEGF and EGFR in CRC appears a promising strategy and warrants evaluation in a prospective study
1 Saltz et al. J Clin Oncol 20072 Cunningham NEngl J Med 2004
Study design CAIRO2
Arm A Arm B
Randomization
CapecitabineOxaliplatin
Bevacizumab
CapecitabineOxaliplatin
BevacizumabCetuximab
Endpoints
• Primary endpoint
progression-free survival
• Secondary endpoints
overall survival, response rate, toxicity,
translational research
Statistical design
• Study was designed to detect a difference in median progression-free survival of 3 months (11m 14m) (HR 0.79), power 80%, =0.05, 2- tailed test
• Stratification parameters- prior adjuvant chemotherapy- serum LDH- number of affected organs- institution
• Histologically proven colorectal cancer
• Advanced disease not amenable to curative surgery
• Measurable disease parameters
• No previous systemic treatment for advanced disease
• Previous adjuvant chemotherapy should be completed 6 months prior to randomization
• Age 18 years
• WHO performance score 0-1
• Adequate hepatic, bone marrow, and renal function
• No therapeutic dose of anticoagulant drugs
• No significant cardiovascular or other disease
Main inclusion criteria
Dose and scheduleall cycles given 3-weekly
Arm ACycle 1-6:
• oxaliplatin 130 mg/m² day 1
• capecitabine 1000 mg/m² b.i.d. day 1-14
• bevacizumab 7.5 mg/kg day 1Cycle ≥ 7:
• capecitabine 1250 mg/m² b.i.d. day 1-14
• bevacizumab 7.5 mg/kg day 1
Arm B
• oxaliplatin, capecitabine, bevacizumab: as in arm A
• cetuximab weekly 250 mg/m² (400 mg/m² 1st dose)
Evaluation of response
• Evaluation of tumor response every 3 cycles (RECIST)
• Evaluation of toxicity prior to each cycle (NCI-CTC criteria, version 3.0)
• Central review was performed of all patient files when death occurred ≤ 30 days of the last administration of study drugs and which was accompanied by any other event than disease progression, irrespective of the causality reported for this event
• All serious adverse events and results of central review were submitted to the IDMC
Accrual
• Participation of 79 Dutch hospitals
• 755 patients were randomized between June 2005 and December 2006
• 736 patients were eligible
• 731 patients received ≥ 1 treatment cycle
• Median duration of follow-up 18.7 months
Arm A Arm B
n = 368 n = 368
Age, median (range) 62 (27-83) 62 (33-80)
Gender * male
female
56%
44%
64%
36%
Prior adjuvant treatment 15% 15%
Serum LDH normal
abnormal
57%
43%
56%
44%
Number of organs affected 1
> 1
45%
55%
44%
56%
WHO performance status 0 59% 66%
* p = 0.035
Baseline characteristics
Arm A Arm B p value
n = 368 n = 368
Median PFS (months)
(HR; 95% CI)
10.7
(9.7-12.5)
9.6
(8.5-10.7)
0.018
(1.21;1.03-1.45)
Median OS (months)
(HR; 95% CI)
20.4
(18.1-26.1)
20.3
(17.9-21.6)
0.21
(1.15;0.93-1.43)
Response rate
(CR + PR)44% 44% 0.88
Disease control rate
(CR + PR + SD)83% 81% 0.39
Efficacy results
Results were confirmed in the subgroup of patients with EGFR+ tumors
0 6 12 18 24 30Months from randomization
0.0
0.2
0.4
0.6
0.8
1.0
Pro
gre
ssio
n f
ree
surv
ival
pro
bab
ility
Arm A (without cetuximab)
Arm B (with cetuximab)
Arm A (without cetuximab) 10.7 months (9.7-12.5) Arm B (with cetuximab) 9.6 months (8.5-10.7)
Progression-free survival
Hazard ratio for progression 1.21
p value 0.018
0 6 12 18 24 30Months from randomization
0.0
0.2
0.4
0.6
0.8
1.0
Ove
rall
surv
ival
pro
bab
ility
Arm A (without cetuximab)
Arm B (with cetuximab)
Overall survival
Arm A (without cetuximab) 20.4 months (18.1-26.1) Arm B (with cetuximab) 20.3 months (17.9-21.6)
Hazard ratio for survival 1.15
p value 0.21
Arm A
n = 366
Arm B
n = 365
p value
All grade 3-4 72% 82% 0.0013
Skin toxicity excluded
(except HFS)72% 75% 0.37
Hematological toxicity 2% 2% 0.99
Diarrhea 19% 26% 0.026
Vomiting 9% 6% 0.20
Hand-foot syndrome 19% 19% 0.98
Neurotoxicity 10% 8% 0.37
GI perforation 0.3% 1.4% 0.10
Venous thromboembolic events
7% 8% 0.66
Toxicity (grade 3-4)
Arm A
n = 366
Arm B
n = 365
All grade acneiform skin reactions 4% 84%*
Grade 3 acneiform skin reactions 0.5% 25%*
All grade nail changes 13% 32%*
Grade 3 nail changes 0.3% 4%*
* p<0.001
Skin toxicity associated with cetuximab
Arm A
n = 368
Arm B
n = 368
Deaths ≤ 30 days of last treatment (any cause)
18 (5%) 18 (5%)
Treatment related 4 (1%) 2 (0.5%)
- GI perforation 3 1
- respiratory insufficiency 1
- neutropenic fever 1
60-day mortality 7 (1.9%) 9 (2.4%)
Mortality
Arm A
n = 366
Arm B
n = 365p value
Number of cycles (range) median 10 (1-39) 9 (1-40) 0.02
mean
12.0 10.6
Reasons for treatment discontinuation
- disease progression including death 53% 50% 0.32
- toxicity 27% 32% 0.22
- secondary metastasectomy 5% 5%
- patient refusal 7% 6%
- other 8% 7%
Treatment characteristics
Arm B (with cetuximab): skin grade 3 (n=104)
Arm B (with cetuximab): skin grade 0-1 (n=109)Arm B (with cetuximab): skin grade 2 (n=152)
PFS according to skin toxicity
Arm B grade 0-1 vs 2 vs 3: p ≤ 0.01
0 6 12 18 24 30months from randomization
0.0
0.2
0.4
0.6
0.8
1.0
pro
gre
ss
ion
su
rviv
al
pro
ba
bil
ity
Arm B (with cetuximab): skin grade 0-1 (N=111)Arm B (with cetuximab): skin grade 2 (n=152)
Arm B (with cetuximab): skin grade 3 (N=102)
Arm A (without cetuximab): overall (N=366)
Arm B (with cetuximab): skin grade 0-1 (n=109)Arm B (with cetuximab): skin grade 2 (n=152)Arm B (with cetuximab): skin grade 3 (n=104)
Arm A (without cetuximab): overall (n=366)
PFS according to skin toxicity
Arm A vs arm B grade 0-1: p < 0.0001
Wildtype
n = 305 (61%)
Mutation
n = 196 (39%)
Arm A 152 (50%) 103 (53%)
Arm B 153 (50%) 93 (47%)
Genotyping by Q-PCR - based assay
No difference in baseline characteristics between patients with KRAS wildtype and mutation (except higher serum LDH in wildtype)
No correlation between KRAS status and cetuximab-related skin toxicity
KRAS genotyping (n=501)
Wildtype
n = 305 (61%)
Mutation
n = 196 (39%)p value
Arm A 152 (50%) 103 (53%)
Arm B 153 (50%) 93 (47%)
Median PFS (months)
Arm A 10.7 12.5 0.92
Arm B 10.5 8.6 0.47
p value 0.10 0.043
KRAS genotyping (n=501)
0 6 12 18 24 30months from randomization
0.0
0.2
0.4
0.6
0.8
1.0
Pro
gre
ssio
n f
ree
surv
ival
pro
bab
ility
KRAS and PFS
Arm A (without cetuximab); KRAS mutantArm B (with cetuximab); KRAS mutantArm A (without cetuximab); KRAS wildtypeArm B (with cetuximab); KRAS wildtype
Wildtype
n = 305 (61%)
Mutation
n = 196 (39%)p value
Arm A 152 (50%) 103 (53%)
Arm B 153 (50%) 93 (47%)
Median OS (months)
Arm A 23.0 24.9 0.90
Arm B 22.2 19.1 0.52
p value 0.49 0.35
KRAS genotyping (n=501)
Conclusions - I
• The addition of cetuximab to capecitabine, oxaliplatin and bevacizumab results in a significantly decreased progression-free survival, without affecting overall survival
• The addition of cetuximab to chemotherapy and bevacizumab results in a significant increase of skin toxicity and diarrhea, however the toxicity is acceptable in both treatment arms
• The grade of cetuximab-related skin toxicity significantly correlates with PFS
• The results of cetuximab are not significantly influenced by KRAS status
• In patients with KRAS mutation the addition of cetuximab to chemotherapy and bevacizumab results in a significant decrease in PFS
• Toxicity as a reason for discontinuation of treatment does not significantly differ between treatment arms, therefore a negative interaction between anti-VEGF and anti-EGFR antibodies cannot be excluded
Conclusions - II
DCCG CAIRO2 study - acknowledgementsInvestigators: C. Smorenburg, Alkmaar; R.Hoekstra, Almelo; C.Rodenburg, Amersfoort; J.van der Hoeven, Amstelveen; A.Cats, M.Geenen, C. van Groeningen, D.Richel, B.de Valk, N.Weijl, Amsterdam; J.Douma, Arnhem; P.Nieboer,Assen; F.Valster, Bergen op Zoom; R.Rietbroek, Beverwijk; A.Ten Tije, Blaricum; O.Loosveld, Breda; D.Kehrer, Capelle a/d IJssel; M.Bos, Delft; Z.Erjavec,Delfzijl; H.Sinnige, C.Knibbeler, Den Bosch; W.Van Deijk, F.Jeurissen, H.Sleeboom Den Haag; H.de Jong, Den Helder; A.Imholz, Deventer; E.Muller, Doetinchem; J. vanden Bosch,Dordrecht; S.Hovenga, Drachten; E.Balk, Ede; G.Creemers, M.Dercksen, Eindhoven; M.Legdeur, Enschede; A.Smals, Geldrop; H.van Halteren, Goes; M. van Hennik, Gorinchem; A.van der Torren, Gouda; G.Hospers, R.de Jong, Groningen; G.de Klerk, Haarlem; P.Zoon, Harderwijk; J.Wals, Heerlen; V.Derleyn, Helmond; H.Dankbaar, Hengelo; S.Luyckx, Hilversum; C. de Swart, Hoofddorp; J.Haasjes, Hoogeveen; W.Meijer, Hoorn; M.Polee, Leeuwarden;M.Tesselaar, Leiden; H.Oosterkamp,Leidschendam; J.Bollen, Lelystad; R.Jansen, Maastricht; P. van der Velden, Middelharnis; P.Slee, Nieuwegein; C.Punt, C.Mandigers, Nijmegen; A.Vos, Oss; M.den Boer, Roermond; D. de Gooyer, Roosendaal; A.Planting, A.vander Gaast, J.Pegels, T.Kok, F.de Jongh, Rotterdam; J.Braun, Schiedam; F.Erdkamp, Sittard; G.Veldhuis, Sneek; A.van Reisen, Terneuzen; C.Kruijtzer, Tiel; H.Roerdink, J. van Riel, Tilburg; D.ten Bokkel Huinink, E.Voest, Utrecht; M. van Diemen, Veghel; G.Vreugdenhil, Veldhoven; M.Werter, Venlo; L.Kerkhofs, Vlissingen; P.Schiphorst, Winterswijk; A.van Bochove, Zaandam; A.Ogilvie, Zoetermeer; A.Honkoop, Zwolle Pathology: J.van Krieken, J.Dijkstra, E.Börger, NijmegenStatisticians: N.Antonini, O.Dalesio, Amsterdam Central Datamanagement: L.Mol, F.van Leeuwen, IKO Nijmegen, Independent Data Monitoring Committee: P. De Mulder †, D.Sleijfer, G.Stoter, Supported by Dutch Cancer Foundation, and unrestricted scientific grants from Merck Serono, Roche, Sanofi-Aventis