Colorectal Cancer Abstracts Oral Session: 6/6/10 Alan P. Venook, M.D. University of California, San Francisco
Jan 21, 2016
Colorectal Cancer Abstracts Oral Session: 6/6/10
Alan P. Venook, M.D.University of California, San Francisco
Colorectal Cancer Abstracts• MACRO
• COIN
• CLOCC
• HORIZON III
Horizon IIIAZD2171 (Cediranib) – VEGF TKI
• FOLFOX / Bevacizumab v. FOLFOX / AZD2171
• Fails to meet primary end-point
• Abstract withdrawn pending results Horizon II
• 5/28/10 press release: Horizon II negative
• To be submitted to ESMO
Colorectal Cancer Acronyms• MACRO – MAintenance in ColoRectal Cancer
• COIN – Cetuximab and OxaliplatIN
• CLOCC – Chemo and LOcal Therapy for CRC
Colorectal Cancer Acronyms• MACRO – MAintenance in ColoRectal Cancer
• COIN – COmplicated to INterpret
• CLOCC – Chemo and LOcal Therapy for CRC
MACRO TRIALTabernero et al
ProgressionRCapecitabineOxaliplatin
Bevacizumabx6 cycles q3w
Bevacizumabuntil progression
N=480
N=239
N=241
CapecitabineOxaliplatin
Bevacizumabx6 cycles q3w
CapecitabineOxaliplatin
Bevacizumabuntil progression
Primary end-point: PFS
MACRO TRIALTabernero et al
ProgressionRCapecitabineOxaliplatin
Bevacizumabx6 cycles q3w
Bevacizumabuntil progression
N=480
N=239
N=241
CapecitabineOxaliplatin
Bevacizumabx6 cycles q3w
CapecitabineOxaliplatin
Bevacizumabuntil progression
Primary end-point: PFS
MACRO TRIALTabernero et al
ProgressionRCapecitabineOxaliplatin
Bevacizumabx6 cycles q3w
Bevacizumabuntil progression
N=480
N=239
N=241
CapecitabineOxaliplatin
Bevacizumabx6 cycles q3w
CapecitabineOxaliplatin
Bevacizumabuntil progression
Primary end-point: PFS
Clinical Question:• Does Bevacizumab monotherapy substitute for
continuing CAPEOX plus Bevacizumab?
More toxicity with CAPEOX
QOL worse with CAPEOX
Is survival compromised?
Is Bevacizumab active as single agent in colorectal cancer ?
Bevacizumab monotherapy
Bergsland et al, ASCO, 2000
FU/LV +/- bev
Bevacizumab monotherapy
Bergsland et al, ASCO, 2000
PR = 10%
Kabbinavar, JCO,2003
Overall Survival ITT
Patients at risk
Follow-up median months (range) 21.1 (0-40) 20.4 (0-38)
Safety: Treatment-related NCI Grade 3-4* AEs
XELOX-BEV N=238 s/a BEV N=238
N % N %
NEUROPATHY SENSORY 59 24.8 18 7.6
DIARRHEA 26 10.9 33 13.9
HAND FOOT SKIN REACTION 29 12.2 16 6.7
FATIGUE 24 10.5 10 4.2
HYPERTENSION 9 3.8 17 7.1
PROTEINURIA 1 0.4 4 1.7
THROMBOSIS 2 0.8 3 1.3
PERFORATION, GI 2 0.8 1 0.4
BLEEDING 1 0.4 1 0.4
OBSTRUCTION/GI . . 1 0.4
CARDIAC ISCHEMIA . . 1 0.4
* Include grade 5
Clinical Question:• Does Bevacizumab monotherapy substitute for
continuing CAPEOX plus Bevacizumab?
Is PFS compromised?
STATISTICS
• Sample Size: 470 patients; 235 per arm– Non-inferiority design – Assuming 10 months as median PFS – Non-inferiority limit of 7.6 m (HR=1.32)– Power = 80%
MACRO Conclusions• This data indicate that a priori specified non-
inferiority cannot be still confirmed, but we can reliably exclude a detriment of larger than 3 weeks in PFS.
• This study suggests that maintenance therapy with single agent BEV is an appropriate option following induction XELOX-BEV in pts with mCRC.
• Further studies evaluating single agent BEV after standard chemotherapy in mCRC are warranted.
Progression Free Survival ITT
LNI: 1.32
Follow-up median months (range) 21.1 (0-40) 20.4 (0-38)
Patients at risk
Progression Free Survival ITT
LNI: 1.32
Follow-up median months (range) 21.1 (0-40) 20.4 (0-38)
Patients at risk
HR: 1.11 (0.89 - 1.37)HR: 1.11 (0.89 - 1.37)
Hazard Ratios
Slope of the survival curve
HR = 1.11
.11 excess hazard over median median of 10 months ≈ 4 weeks
HR = 1.11 (0.89 – 1.37) – 95% confidence intervals
as much as .37 hazard over medianmedian of 10 months ≈ 3.7 months
Conclusions• This data indicate that a priori specified non-
inferiority cannot be still confirmed, but we can reliably exclude a detriment of larger than 3 weeks in PFS.
My Conclusions• This data indicate that a priori specified non-
inferiority cannot be still confirmed, but we can reliably exclude a detriment of larger than 3 weeks in PFS.
• However, given overall survival results and the QOL associated with the control arm – which reflects excessive oxaliplatin neurotoxicity – PFS less relevant in this analysis
monthsmonths which definitely does not prove non-inferiority
which definitely does not prove non-inferiority
COIN Trial design
.
5FU or capecitabineoxaliplatin
Arm A
CONTINUOUS CT until progression, toxicity or patient choice
5FU or capecitabineoxaliplatin
cetuximab
Arm B
CONTINUOUS CT until progression, toxicity or patient choice
815
815
Maughan et al
For all patientsIncreased non-haematological toxicityNo change in OS or PFS
For KRASwt patientsNo change in OS or PFSIncreased response rate
COIN Trial design
.
5FU or capecitabineoxaliplatin
Arm A
CONTINUOUS CT until progression, toxicity or patient choice
5FU or capecitabineoxaliplatin
cetuximab
Arm B
CONTINUOUS CT until progression, toxicity or patient choice
815
815
Maughan et al
For all patientsIncreased non-haematological toxicityNo change in OS or PFS
For KRASwt patientsNo change in OS or PFSIncreased response rate
Capecitabine / OxalIplatin
Negative
Capecitabine / OxalIplatin
Negative
COIN conclusions• Subgroup analyses suggest that there may be a benefit for
cetuximab in combination with oxaliplatin chemotherapy in patients with– KRAS wildtype tumours, – Limited metastatic disease (0/1 metastatic sites), – Used in combination with infusional 5FU and oxaliplatin
• The differential benefit for choice of fluoropyrimidine and distribution of disease requires validation from other datasets
Mutations in Kras, Nras, Braf: distribution and prognostic significance
BRAF mutation All patientsAny mutationKRAS mutation
KRAS wild-typeAll wild-type
Mutation status:
06
12
Med
ian
PFS
(m
on
ths) Arm A Arm B
06
12
18
Med
ian
OS
(m
on
ths)
57340
268815
367289
45366
297815
362292
01
02
03
04
0
2-y
ear
OS
(%
)
N:
Prognostic effect of mutational status
“All-wt”n=581 (44%)
KRAS-mutn=565 (43%)
NRAS-mutn=50 (4%)
BRAF-mutn=102 (8%)
Totaln=1316 (81%)
554
11
39
102
Population N Arm A Arm B
ITT 1630 815 815
Assessed for mutations
1316 648 668
of which:- KRAS mutation- NRAS mutation- BRAF mutation
565 (43%)50 (4%)
102 (8%)
2681857
2973245
KRAS wt 729 (55%)
367 306
KRAS/NRAS/BRAF-wt“All wild-type”
581 (44%)
289 292
BRAF mutation All patientsAny mutationKRAS mutation
KRAS wild-typeAll wild-type
Mutation status:
06
12
Med
ian
PFS
(m
on
ths) Arm A Arm B
06
12
18
Med
ian
OS
(m
on
ths)
57340
268815
367289
45366
297815
362292
01
02
03
04
0
2-y
ear
OS
(%
)
N:
Prognostic effect of mutational status
BRAF mutation: poorest prognosis
KRAS wt: superior prognosis compared to KRAS mut
Overall Survival:by mutation status: KRAS, NRAS, BRAF
Arm A Arm B Diff.
Median survival:(mo) 20.1 19.9 -0.20
2-year survival rates
40.0% 38.8%
-1.24%
Arm A Arm B Diff.
Median survival: (mo) 14.4 12.7 -1.64
2-year survival rates 21.2% 25.5%
+4.29%
All wild type Any mutation
0.00
0.25
0.50
0.75
1.00
Sur
viva
l
0 6 12 18 24 30 36 42Time (months)
Arm A (OxFp)
Arm B (OxFp + cetux)
HR point estimate = 1.01995% CI = (0.86, 1.20)Χ2 = 0.03; p = 0.86
0 6 12 18 24 30 36 42Time (months)
Arm A (OxFp)
Arm B (OxFp + cetux)
HR point estimate = 1.00495% CI = (0.87, 1.15)Χ2 = 0.00; p = 0.96
Last Conclusion, and not in written abstract• (0/1 metastatic sites),
• Strong prognostic effect of KRAS, BRAF and NRAS mutation status independent of the use of cetuximab
consistent with BRAF data, not conventional wisdom for KRAS
consistent with BRAF data, not conventional wisdom for KRAS
Cetuximab v. BSC 3rd-line:OS by KRAS status in BSC arms
0
0.2
0.4
0.6
0.8
1
0 2 4 6 8 10 12 14 16 18
Time from Randomisation (Months)
Pro
po
rtio
n A
liv
e
Mutated
Wild Type
MutatedWild Type
83 69 42 28 20 13 11 7113 92 69 36 24 17 12 5
HR HR 1.01 1.01 95% CI (0.74,1.37) 95% CI (0.74,1.37)
Log rank p-value: Log rank p-value: 0.970.97
KRAS statusKRAS status MS MS (months)(months)
95% CI95% CI
MutatedMutated 4.64.6 3.6 – 5.53.6 – 5.5
Wild-TypeWild-Type 4.84.8 4.2 – 5.54.2 – 5.5
Karapetis et al, NEJM, 2008
KRAS mutation: Predictive and/or prognostic ?
PREDICTIVE – response to therapy• KRAS mutation predicts non-response to EGF-R antibodies
PROGNOSTIC – outcome independent of treatment• NO:
– 3rd line Cetuximab / Panitumumab monotherapy trials– CAIRO-2 – 1st line oxaliplatin / bevacizumab +/- cetuximab stage IV– PACCE – 1st line chemo / bevacizumab +/- panitumumab stage IV
• YES:– RASCAL – stage III colon cancer– FOCUS – 1st line sequential treatment stage IV– COIN – 1st line oxaliplatin / cetuximab stage IV
• MAYBE:– CRYSTAL – 1st line FOLFIRI / cetuximab stage IV
KRAS mutation: Predictive and/or prognostic ?
PREDICTIVE – response to therapy• KRAS mutation predicts non-response to EGF-R antibodies
PROGNOSTIC – outcome independent of treatment• NO:
– 3rd line Cetuximab / Panitumumab monotherapy trials– CAIRO-2 – 1st line oxaliplatin / bevacizumab +/- cetuximab stage IV– PACCE – 1st line chemo / bevacizumab +/- panitumumab stage IV
• YES:– RASCAL – stage III colon cancer– FOCUS – 1st line sequential treatment stage IV– COIN –
• MAYBE:– CRYSTAL – 1st line FOLFIRI / cetuximab stage IV
N0147 – later in this sessionN0147 – later in this session
My conclusions
• KRAS not prognostic 3rd line– May reflect selection of patients:
• KRAS mutant patients who proceed to a 3rd line study may be a subset not representative of the KRAS mutant patients in general
• If KRAS status is prognostic, then– Current studies may need to be stratified for
KRAS status and analyzed for KRAS prognostic status
– Past study results may be in question because of the uncertain KRAS distribution
CALGB/SWOG 80405
Untreatedadvancedor mCRCN =1142
Bevacizumabfollowed by
FOLFOX or FOLFIRIq 2 wks
Cetuximabfollowed by
FOLFOX or FOLFIRIq 2 wks
Accrual 6/2/10 = 606
Screenfor
eligibility
Sendtumortissue
block toSWOGPCO
RandomizePatients
w/Wild type
K-ras tumor
RegisterPatient
CLOCC 40004 Ruers et al
Randomi z e
RF + chemotherapy ± additional resection
Patients with unresectable CRC liver metastases
N=152Chemotherapy good response resectionN = 119
Local Therapy for Liver Tumors- BUT DO THEY IMPROVE OUTCOMES ?
• Hepatic Intra-arterial chemotherapy• Trans-arterial chemoembolization• Isolated hepatic perfusion• Radio-frequency Ablation• Ablation
– Cryo, microwave, ultrasound, irreversible electroporation
• External radiotherapy• Internal radiation
Key details
Unresectable “incurable” liver metastases
No extrahepatic disease Prior chemotherapy allowed as long as at least SD
Primary objective:
To demonstrate a 30-month OS >38% in RF + chemotherapy arm Fleming design, power of 90% and 1 sided type I error of 10%
07
Primary end point: 30 months Overall Survival
TreatmentPatients
(N)
ObservedEvents
(O)Hazard Ratio
(95% CI)P-Value
(Log-Rank)Median (95% CI)
(Years)
% at 30 months (95% CI)
Chemo 59 39 1.00 0.2176 3.38 (2.46, 4.18)
58.56 (44.82, 69.99)
RF +Chemo
60 31 0.74 (0.46, 1.19)
3.78 (2.76, N) 63.83(50.10,74,71)
Primary end point: 30-month OS in RFA + chemo >38% met in BOTH ARMS
My Conclusions
• RFA + chemotherapy with FOLFOX (+ bevacizumab) is safe and feasible
• Patients with liver-only metastases do appreciably better than patients with other sites of disease
• It is tough / impossible to accrue to studies randomizing patients with liver-only disease to this kind of intervention because it is offered in the community
• This important clinical and cost-economics question is unlikely to ever be addressed in a randomized trial
Colorectal Cancer Abstracts: Comments and Conclusions
• Well-done and well-presented• Europeans get studies done• Americans: studies are under-powered• Important clinical questions
– Maintenance Bevacizumab– Patient selection – Role for RFA debulking
• Did not get conclusive answers because…
Colorectal Cancer Abstracts: Comments and Conclusions• What may be new:
– KRAS MAY be prognostic
• What we knew and still know:– Pts with liver-only metastases do well with or
without RFA and RFA will continue to be done
• What we still do not know:– Bevacizumab as maintenance– What the message is from the COIN trial
• What we need to do now:– Go back to the drawing board