Adjuvant Matters

Adjuvant MattersAdjuvant Matters

Richard M Goldberg MDRichard M Goldberg MDUNC Lineberger Comprehensive UNC Lineberger Comprehensive

Cancer CenterCancer CenterChapel Hill, NC Chapel Hill, NC

DisclosuresDisclosures

I have been a paid consultant to I have been a paid consultant to sanofi-aventissanofi-aventis GenentechGenentech PfizerPfizer

MembershipsMemberships NSABP (Wolmark, Allegra abstracts) NSABP (Wolmark, Allegra abstracts) ACCENT (de Gramont abstract)ACCENT (de Gramont abstract)

Coauthor Coauthor Ribic NEJM paper upon which some of the Ribic NEJM paper upon which some of the

presentation by Sargent is basedpresentation by Sargent is based

NSABP Protocol C-07:A Phase III Trial Comparing FU/LV to FU/LV + Oxaliplatin in Stage II or III Colon Cancer:

Survival Data

Wolmark, Wieand, Kuebler, Colangelo, O'Connell, Yothersand the NSABP Investigators

Questions Questions

AnsweredAnswered Does oxaliplatin add to PFS and OS?Does oxaliplatin add to PFS and OS? Does 5-FU schedule matter?Does 5-FU schedule matter? Is 5 years of follow-up adequate for OS?Is 5 years of follow-up adequate for OS?

UnansweredUnanswered What’s the optimal adjuvant therapy duration?What’s the optimal adjuvant therapy duration? How much total oxaliplatin is optimal?How much total oxaliplatin is optimal? How do we best manage Stage II patients? How do we best manage Stage II patients?

Stage ll + lll

FLOXFU/LV

Randomize

Critical StatisticsCritical Statistics

5.5 year median follow-up5.5 year median follow-up Outcomes better than projectedOutcomes better than projected

Expected # deaths 700Expected # deaths 700 Actual # deaths 560Actual # deaths 560 Consequent reduction in powerConsequent reduction in power

% Stage II: % Stage II: C-07 28% C-07 28% MOSAIC 40%MOSAIC 40%

p = 0.002HR: 0.81 [0.70 – 0.93]

19% risk reduction

C-07 DFS

3y 5yFLOX 76.1% 69.4%FULV 71.5% 64.2% Δ 4.6% 5.2%

D(n) 5y 6yFLOX 259 80.3% 77.7%FULV 301 78.3% 73.5% Δ 42 2.0% 4.2% p = 0.06HR: 0.85 [0.72 – 1.01]

15% risk reduction

C-07 Survival

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1

0 1 2 3 4 5

Years

Median FLOX 17.6 FULV 22.2

P = 0.02HR: 1.24

C-07 Survival after Recurrence

Questions Questions

AnsweredAnswered Does oxaliplatin add to PFS and OS? Does oxaliplatin add to PFS and OS? YesYes Does 5-FU schedule matter? Does 5-FU schedule matter? NoNo Is 5 years of follow-up adequate for OS? Is 5 years of follow-up adequate for OS? NoNo

UnansweredUnanswered What’s the optimal adjuvant therapy duration? What’s the optimal adjuvant therapy duration? 3 3

versus 6 versus other?versus 6 versus other? How much total oxaliplatin is optimal? How much total oxaliplatin is optimal? May differ in May differ in

different patients different patients How do we best manage Stage II patients? How do we best manage Stage II patients?

TBDTBD No treatment, 5-FU alone, or FOLFOXNo treatment, 5-FU alone, or FOLFOX

Initial Safety Report of NSABP C-08 Initial Safety Report of NSABP C-08

Allegra, Yothers, Sharif, O’Connell, Allegra, Yothers, Sharif, O’Connell, Wolmark and the NSABP InvestigatorsWolmark and the NSABP Investigators

NSABP C-08 SchemaNSABP C-08 SchemaStage II or III Colon

Cancer

StratificationNumber of + Lymph Nodes

Randomization

mFOLFOX6 mFOLFOX6 +

Bevacizumab

Disease-free survival:Disease-free survival: primary endpoint primary endpoint

QuestionsQuestions

AnsweredAnswered Is the frequency of early deaths different?Is the frequency of early deaths different? What are the relative toxicities of FOLFOX +/- What are the relative toxicities of FOLFOX +/-

bevacizumab?bevacizumab? Were there any surprises?Were there any surprises?

UnansweredUnanswered Does bevacizumab add to FOLFOX in the Does bevacizumab add to FOLFOX in the

adjuvant setting?adjuvant setting? Does bevacizumab have any long term Does bevacizumab have any long term

toxicity (or benefit)?toxicity (or benefit)?

Early Mortality with Early Mortality with Adjuvant RegimensAdjuvant Regimens

C-08 within 18 months of randomizationC-08 within 18 months of randomization

• FOLFOX 1.33%FOLFOX 1.33%• FOLFOX + bev 1.42%FOLFOX + bev 1.42%

MOSAIC data within 7 months of randomizationMOSAIC data within 7 months of randomization• LV5FU2 0.5%LV5FU2 0.5%• FOLFOX 0.5%FOLFOX 0.5%

C89803 data within 6 months of randomizationC89803 data within 6 months of randomization• 5-FU/LV 1%5-FU/LV 1%• IFL 2.8%IFL 2.8%

Toxicities (Grade 3+) Toxicities (Grade 3+) Significantly Increased with Significantly Increased with

Bevacizumab (%)Bevacizumab (%)

mFOLFOX6mFOLFOX6 mFOLFOX6mFOLFOX6+ Bev+ Bev p-valuep-value

HypertensionHypertension 1.81.8 1212(5 pts Gr 4)(5 pts Gr 4)

<0.0001<0.0001

Any PainAny Pain 6.36.3 11.111.1 <0.0001<0.0001

ProteinuriaProteinuria 0.80.8 2.72.7 <0.001<0.001

Wound ComplicationsWound Complications 0.30.3 1.71.7(all Gr 3)(all Gr 3)

<0.001<0.001

mFOLFOX6mFOLFOX6(%)(%)

mFOLFOX6mFOLFOX6+ Bev (%)+ Bev (%)

p-valuep-value

Wound Wound ComplicationsComplications

0.250.25 1.111.11 0.010.01

HypertensionHypertension 0.740.74 5.855.85 <0.0001<0.0001

Sensory Sensory NeuropathyNeuropathy

(Grade 2+)(Grade 2+)

26.126.1 32.432.4 <0.001<0.001

DepressionDepression 1.321.32 2.932.93 <0.01<0.01

DizzinessDizziness 0.660.66 1.581.58 0.040.04

ProteinuriaProteinuria 0.20.2 1.61.6 <0.001<0.001

Pain - AnyPain - Any 2.12.1 4.74.7 <0.001<0.001

Toxicities (Grade 3+) After ChemotherapyToxicities (Grade 3+) After Chemotherapy

QuestionsQuestions AnsweredAnswered

Is the frequency of early deaths different? Is the frequency of early deaths different? NoNo What are the relative toxicities of FOLFOX +/- What are the relative toxicities of FOLFOX +/-

bevacizumab? bevacizumab? BP, pain, wounds, proteinuriaBP, pain, wounds, proteinuria Were their any surprises? Were their any surprises? Pain, a bit more oxaliplatin Pain, a bit more oxaliplatin

delivered, no major increase in wound complications delivered, no major increase in wound complications UnansweredUnanswered

Does bevacizumab add to FOLFOX in the adjuvant Does bevacizumab add to FOLFOX in the adjuvant setting? setting? No dataNo data

• 28 months median time on study28 months median time on study• Efficacy analysis: 592 events or 4 years after last entry Efficacy analysis: 592 events or 4 years after last entry

(DMC evaluation every 6 months, latest October 2010)(DMC evaluation every 6 months, latest October 2010) Does bevacizumab have any long term toxicity (or Does bevacizumab have any long term toxicity (or

benefit)? benefit)? No dataNo data

Association between 3 year DFS and OS is Association between 3 year DFS and OS is delayed with improved survival after recurrence in delayed with improved survival after recurrence in patients receiving cytotoxic adjuvant therapy for patients receiving cytotoxic adjuvant therapy for colon cancer: Findings from the 20,898 patient colon cancer: Findings from the 20,898 patient

ACCENT datasetACCENT dataset

de Gramont for the de Gramont for the ACCENT collaborative groupACCENT collaborative group

QuestionsQuestions

AnsweredAnswered Do we need to extend adjuvant trial follow-up Do we need to extend adjuvant trial follow-up

beyond 5 years?beyond 5 years? Is a survival advantage that is apparent after Is a survival advantage that is apparent after

5 years still meaningful?5 years still meaningful? Should trial design be continuously Should trial design be continuously

reevaluated?reevaluated? UnansweredUnanswered

Do we need to change surveillance plans?Do we need to change surveillance plans?

MOSAIC Update: OS with 6 Years MOSAIC Update: OS with 6 Years Minimum Follow-upMinimum Follow-up

FOLFOX4 stage II

LV5FU2 stage II

FOLFOX4 stage III

LV5FU2 stage III

Overall survival (months)

Pro

bab

ilit

y

1.0

0.8

0.6

0.4

0.2

0

0.9

0.7

0.5

0.3

0.1

0 6 12 18 24 6030 36 42 48 54 66 9672 78 84 90

HR [95% CI]

Stage II 1.00 [0.71–1.42]

Stage III 0.80 [0.66–0.98]

0.1%

4.4%

p=0.996

p=0.029

ASCO 2007

0.5

0.6

0.7

0.8

0.9

1

1.1

1.2

1.3

0.5 0.6 0.7 0.8 0.9 1 1.1 1.2 1.3

HR for 3 Year DFS

HR

fo

r 5

Ye

ar

OS

R2 = 0.80

ACCENT: 3yr DFS vs 5yr OSACCENT: 3yr DFS vs 5yr OS

May 2004: ODAC recommends3-yr DFS as new regulatory endpointfor FULL approval in adjuvant colon cancer

0.5

0.6

0.7

0.8

0.9

1

1.1

1.2

1.3

0.5 0.6 0.7 0.8 0.9 1 1.1 1.2 1.3

HR for 3 Year DFS

HR

fo

r 7

Ye

ar

OS

Hypothetical – 3yr DFS v. 7yr OSHypothetical – 3yr DFS v. 7yr OS

R2 = 0.75

Why Was 6 Years Required for Why Was 6 Years Required for MOSAIC to Become Positive for OS?MOSAIC to Become Positive for OS?

Previous analyses with 5FU/LV showed Previous analyses with 5FU/LV showed excellent association between 3 yr DFS & 5 excellent association between 3 yr DFS & 5 yr OSyr OS ACCENT: Median time from recurrence ACCENT: Median time from recurrence

to death: 12 monthsto death: 12 months MOSAIC: Median ~ 24 monthsMOSAIC: Median ~ 24 months

• Improved imaging to detect recurrence Improved imaging to detect recurrence earlierearlier

• Enhanced treatment post-recurrenceEnhanced treatment post-recurrence• Secondary resections in selected patientsSecondary resections in selected patients

Impact of longer survival following Impact of longer survival following recurrence on clinical trialsrecurrence on clinical trials

Longer follow-up for OS required to Longer follow-up for OS required to observe benefitobserve benefit improved post-recurrence treatmentsimproved post-recurrence treatments

DFS even a more important endpointDFS even a more important endpoint Treatments that delay rather than Treatments that delay rather than

prevent recurrence may send prevent recurrence may send misleading DFS signalsmisleading DFS signals

QuestionsQuestions

AnsweredAnswered Do we need to extend adjuvant trial follow-up beyond Do we need to extend adjuvant trial follow-up beyond

5 years? 5 years? YesYes Is a survival advantage only apparent after 5 years Is a survival advantage only apparent after 5 years

still meaningful? still meaningful? AbsolutelyAbsolutely Should trial design be continuously reevaluated? Should trial design be continuously reevaluated?

Absolutely, by clinician & statistician teamsAbsolutely, by clinician & statistician teams

UnansweredUnanswered Do we need to change surveillance plans? Do we need to change surveillance plans? ProbablyProbably

Deficient Mismatch Repair as a Deficient Mismatch Repair as a Predictive Marker for Lack of Predictive Marker for Lack of

Benefit from 5-FU based Benefit from 5-FU based Chemotherapy in Adjuvant Colon Chemotherapy in Adjuvant Colon

CancerCancer

Sargent, Marsoni, Thibodeau, Labianca, Sargent, Marsoni, Thibodeau, Labianca, Hamilton, Torri, Monges, Ribic, Grothey, Hamilton, Torri, Monges, Ribic, Grothey,

GallingerGallinger

QuestionsQuestions

AnsweredAnswered Should MMR testing be considered prior to 5-Should MMR testing be considered prior to 5-

FU based Rx for stage II pts?FU based Rx for stage II pts?

UnansweredUnanswered Should MMR testing be considered prior to 5-Should MMR testing be considered prior to 5-

FU based Rx for stage III pts?FU based Rx for stage III pts? Should other agents be used in MSI-H pts?Should other agents be used in MSI-H pts? Why are MSI-H different then MSS tumors?Why are MSI-H different then MSS tumors?

BiologyBiology

15% of CRC pts have MSI-H tumors15% of CRC pts have MSI-H tumors Most hypermethylation, not mutationMost hypermethylation, not mutation

MSI-H cells have a growth advantage in MSI-H cells have a growth advantage in the presence of 5-FU the presence of 5-FU Carrethers, Gastro, 1999Carrethers, Gastro, 1999 MMR cells unable to bind 5-FU modified DNAMMR cells unable to bind 5-FU modified DNA

MSI-H cell linesMSI-H cell lines are more sensitive to irinotecan are more sensitive to irinotecan

Bras-Goncalves, BJC, 2000Bras-Goncalves, BJC, 2000 are not resistant to oxaliplatin are not resistant to oxaliplatin

Sergent, Canc Chemo Pharmacol 2002Sergent, Canc Chemo Pharmacol 2002

Pooled data (N=1027)Pooled data (N=1027)TrialTrial Treatment Treatment NN % Stage II% Stage II % dMMR% dMMR

784852784852 5FU/LEV5FU/LEV 117117 30%30% 14%14%

INT 0035INT 0035 5FU/LEV5FU/LEV 215215 50%50% 18%18%

874651874651 5FU/LV5FU/LV 6666 19%19% 12%12%

GIVIOGIVIO 5FU/LV5FU/LV 183183 52%52% 16%16%

FFCDFFCD 5FU/LV5FU/LV 154154 66%66% 19%19%

NCICNCIC 5FU/LV5FU/LV 292292 61%61% 15%15%

TotalTotal 10271027 52%52% 16%16%

DFS By MMR Status, DFS By MMR Status, Pooled DataPooled Data

01 02 03 04 05 06 07 08 09 0

1 0 0

0 1 2 3 4 5Y e a rs

% D

ise

as

e F

ree

0102030405060708090

100

0 1 2 3 4 5Years

% D

isea

se F

ree

HR: 0.79 (0.49-1.25)p=0.30 HR: 0.51 (0.29-0.89)

p=0.009

Treated (N=512) Untreated (N=515)

dMMR 70%pMMR 67%

5 yr DFS

dMMR 80%pMMR 56%

5 yr DFS

Overall Survival By Treatment, Overall Survival By Treatment, Stage II dMMR PatientsStage II dMMR Patients

0

10

20

30

40

50

60

70

80

90

100

0 1 2 3 4 5 6 7 8Years

% A

live

HR: 3.15 (1.07-9.29)p=0.03

N = 55

N = 47

Untreated 93%Treated 75%

5 yr OS

P-value = 0.014 for treatment by MMR status interaction

ConclusionsConclusions dMMR validated as a prognostic marker in dMMR validated as a prognostic marker in

untreated patientsuntreated patients

No suggestion of benefit from 5-FU based No suggestion of benefit from 5-FU based treatment in dMMR patientstreatment in dMMR patients

Significant OS decrement to 5-FU based Significant OS decrement to 5-FU based treatment in stage II patientstreatment in stage II patients

Take Home MessageTake Home Message

When considering a Stage II patient for When considering a Stage II patient for

5-FU-based therapy, 5-FU-based therapy,

mismatch repair status, mismatch repair status,

by MSI or IHC, by MSI or IHC,

should be used to determine should be used to determine

whom not to treatwhom not to treat

CALGB 89803: CALGB 89803: 5-FU/ LV +/- Irinotecan5-FU/ LV +/- Irinotecan

854/1264 tumors evaluated for MSI854/1264 tumors evaluated for MSI Patients with samples no different from overall Patients with samples no different from overall

population population 153 (18%) MSI-H by DNA microsatellite 153 (18%) MSI-H by DNA microsatellite

analysisanalysis PFS (p=0.04) advantage and OS trend PFS (p=0.04) advantage and OS trend

(p=0.17) for IFL over FL treated MSI-H (p=0.17) for IFL over FL treated MSI-H patientspatients

No PFS or OS advantage for MSS patientsNo PFS or OS advantage for MSS patients Bertagnolli, Bertagnolli,

submittedsubmitted

p= 0.0391 p = 0.1736

p = 0.5111 p = 0.6681

Disease free survival Overall survival

MSI-Htumors

Non-MSI-Htumors

QuestionsQuestions AnsweredAnswered

Should MMR testing be considered prior to 5-FU Should MMR testing be considered prior to 5-FU based Rx for stage II pts? based Rx for stage II pts? YesYes

UnansweredUnanswered Should MMR testing be considered prior to 5-FU Should MMR testing be considered prior to 5-FU

based Rx for stage III pts? based Rx for stage III pts? UnclearUnclear Should other agents be used in MSI-H pts? Should other agents be used in MSI-H pts?

Needs verificationNeeds verification• Irinotecan and oxaliplatin cell line dataIrinotecan and oxaliplatin cell line data• CALGB clinical dataCALGB clinical data

Why are MSI-H different then MSS tumors? Why are MSI-H different then MSS tumors? TBDTBD

ConclusionsConclusionsAdjuvant MattersAdjuvant Matters

5-FU/Oxaliplatin standard for adjuvant Rx of 5-FU/Oxaliplatin standard for adjuvant Rx of Stage III colon cancerStage III colon cancer Should be considered in high-risk stage IIShould be considered in high-risk stage II No advantage over 5-FU/LV in pooled stage IINo advantage over 5-FU/LV in pooled stage II

Bevacizumab is not standard adjuvant Rx Bevacizumab is not standard adjuvant Rx Adjuvant trials need to be bigger and longerAdjuvant trials need to be bigger and longer MSI status appears relevent to likelihood of MSI status appears relevent to likelihood of

benefit from 5-FU and possibly other agentsbenefit from 5-FU and possibly other agents At last we have useful markers of heterogeneityAt last we have useful markers of heterogeneity

K-ras and MSI appear to be prognostic and predictive K-ras and MSI appear to be prognostic and predictive