Edith A. Perez, MDDirector, Cancer Clinical Study Unit

Mayo ClinicJacksonville, Florida

Welcome and Introductions

John F. Forbes, MB, BS, MS, FRACSProfessor of Surgical Oncology

University of NewcastleDirector, Department of Surgical Oncology

Newcastle Mater Misericordiae HospitalNewcastle, Australia

The Early Use of Adjuvant Aromatase Inhibitors for Early Breast Cancer:

New Contributions from the BIG 1-98 Letrozole Trial

44

Outline

• Trial Design

• Statistical Analyses

• Population

• Efficacy Endpoints

• Subgroups

• Safety

• Conclusions

• Perspective

55

Argentina 123 New Zealand 157

Australia 667 Peru 51

Belgium 634 Poland 277

Brazil 17 Portugal 64

Canada 20 Russia 240

Chile 22 Slovenia 15

Czech Rep. 109 South Africa 187

Denmark 1396 Spain 70

France 1016 Sweden 64

Germany 113 Switzerland 611

Hungary 334 Turkey 54

Iceland 6 United Kingdom 401

Italy 1285 Uruguay 1

Netherlands 94 TOTAL 8028

BIG 1-98 Worldwide Collaborative8028 patients enrolled March 1998-May 2003

66

BIG 1-98 Design

Tamoxifen

Letrozole

Letrozole

Letrozole Tamoxifen

RANDOMIZE

0 2 5YEARS

A

B

C

D

• Compares Letrozole versus Tamoxifen• Letrozole: Arms B and D • Tamoxifen: Arms A and C• Excludes events and FU beyond switch for C & D

Tamoxifen

77

BIG 1-98

New since St. Gallen (January 2005)- Medical review of all cerebrovascular, cardiac,

unclear AEs (538 cases) and all deaths without prior cancer event (93 cases)

- Overall survival outcome by subgroups

- Identification of myalgia and arthralgia AEs

Still to come- Central review of ER, PgR, Her-2

- Update of safety and efficacy

- Results of sequential treatment comparisons

88

Primary Core Analysis

8028 Randomized

8010 Primary Core Analysis

4007 T4003 L versus

18 withdrew consent (no treatment / FU)

133 (1.66%) ineligible cases included in primary core analysis

Median Follow-Up=25.8 months

99

Patient/Tumor Characteristics

Letrozole Tamoxifen

Median age 61 61

Tumor size > 2 cm 36.5% 37.7%

Node positive 41.5% 41.2%

Chemotherapy given 25.3% 25.3%

ER+ / PgR+ 63.5% 62.7%

ER+ / PgR- 20.2% 20.5%

ER+ / PgR unk 14.5% 14.3%

Receptor positivity was a study requirement:99.8% of patients had receptor positive tumors

1010

Primary End Point: DFS

Time from randomization to first of: Invasive recurrence in

- Ipsilateral breast

- Chest wall

- Regional site (internal mammary/axilla)

- Distant site (including ipsi supraclavicular)

Contralateral breast (invasive)

Second (non breast) malignancy

Death without prior cancer event

1111

Secondary End Points

Overall survival (OS)

Systemic disease-free survival (SDFS)*

Distant disease-free survival (DDFS)**

Safety

* Excluding locoregional and contralateral events** Excluding locoregional and 2nd non-breast cancer

1212

T

0

20

40

60

80

100

0 1 2 3 4 5

Per

cen

t A

live

and

Dis

ease

-Fre

e

Years from Randomization

Disease-Free Survival

L

97.797.6

YearlyDFS %

95.193.4

90.589.0

86.884.6

84.081.4

No. at Risk

38923896

29642926

12611238

892866

40034007

567544

N HR (95% CI) p

8010 0.81 (.70-.93) 0.003

Events

779

1313

Years from Randomization

0 2 3 4 51

0

10

5

15

20

Pro

port

ion

Fa

ilure

(%

)

L

T

Cumulative IncidenceBreast Cancer Event

13.6%

10.2%

8.1%

6.2%

5-year diff (L-T) = -3.4% (S.E. 1.2)Cum incidence P=0.0002

1414

Treatment Failures

Letrozole Tamoxifen P

First Failure Sites (DFS events) 8.8% 10.7% 0.003

Local 0.5% 0.9% 0.034

Contralateral Breast (invasive) 0.4% 0.7% 0.092

Regional* 0.3% 0.3% 0.842

Distant 4.4% 5.8% 0.005

Second (non breast) malignancy 1.7% 2.0% 0.288

Death without cancer event 1.4% 0.9% 0.077

Deaths 4.1% 4.8% 0.155

Systemic Failures** 8.1% 9.6% 0.017

*Regional includes axilla or internal mammary**SDFS ignores local and contralateral events

1515

Deaths

Letrozole Tamoxifen

Patients 4003 4007

Deaths 166 192

Deaths following cancer event 111 154

Deaths w/o prior cancer event 55 38

- Cerebro-vascular accident 7 1

- Venous thromboembolic 2 2

- Cardiac 13 6

- Sudden death (cause unk) 10 10

- Other 23 19

1616

Protocol Endpoints

DFS

OS

SDFS

Favors L Favors T

0.81

0.86

0.83

1.00.5 0.75 1.33 2.0

Hazard Ratio (L:T)

1717

Other Endpoints

DFS

OS

SDFS

Time to recurrence

DFS (w/o 2nd malignancy)

Favors L Favors T

0.81

0.86

0.83

0.79

0.73

1.00.5 0.75 1.33 2.0

Hazard Ratio (L:T)

Time to distant recurrence

0.72

1818

Sub group Analyses

• Subgroup analyses should concentrate on:

differences from the average overall treatment

effect (via tests of heterogeneity or interaction)

• It is inappropriate to assess the effects of

treatment on a single subgroup by examination

of the 95% CI for that subgroup.

Cuzick J 1982; Lancet 2005 365:1308

1919

Sub group Analyses

Two types of error can occur

1. Attribution of an effect to a subgroup

when there is no overall effect and no evidence for heterogeneity (more common)

2. To claim a lack of effect in a subgroup

when the overall effect is significant

Cuzick J 1982; Lancet 2005 365: 1308

2020

Sub group Analyses

• Confidence intervals in subgroups are always wider than

those for the main effect because of smaller numbers.

• If the interval for a subgroup crosses the no effect point,

this is widely misinterpreted as a lack of effect in the

subgroup even when the overall effect is significant.

• The correct approach is to determine whether the effect

size for different subgroups varies significantly from the

main effect by a test for heterogeneity.

Cuzick J 1982; Lancet 2005 365: 1308

2121

Subgroups - DFS

Favors L Favors T

1.00.5 0.75 1.33 2.0

Hazard Ratio (L:T)

CT given (n=2024)

CT not given (n=5986)

0.70

0.85

N-positive (n=3311)

N-negative (n=4174)

0.71

0.99

ER+ / PgR+ (n=5055)*

ER+ / PgR- (n=1631)*

0.84

0.83

* Based on local assessment

2222

Subgroup - OS

0.76

0.90

Favors L Favors T

1.00.5 0.75 1.33 2.0

Hazard Ratio (L:T)

CT given (n=2024)

CT not given (n=5986)

0.82

0.88

N-positive (n=3311)

N-negative (n=4174)

1.00

0.79

ER+ / PgR+ (n=5055)*

ER+ / PgR- (n=1631)*

* Based on local assessment

2323

Summary of Efficacy

• Letrozole significantly decreased overall risk of recurrence (19% P=0.003)

• Letrozole significantly reduced the risk of distant metastases (27% P=0.0012)

• Letrozole was associated with a non significant decreased risk of death (14% P=0.16)

• The results are consistent with a similar effect in all subgroups examined

2424

Adverse Events, Any Grade

38.0

6.6

3.5

43.5

20.3

6.4

5.7

4.1

1.01.0

3.8

4.0

6.1

12.3

19.1

16.2

1.5

3.3

13.9

33.5

0 25 50

Night Sweats

Hot flushes

Hyperchol*

Joint

Muscle

Vaginal bleeding

Bone fracture

Cardiac

Thromboembolic

CVA/TIA

Percent of Patients

Letrozole

Tamoxifen

*Grade 1: 35.1% L, 17.3% T; Grade 2+: 8.5% L, 1.9% T Serial cholesterol levels are being reviewed.

2525

Cardiovascular Events, Grade 3-5

Letrozole Tamoxifen

Patients 3975 3988 P

CVA/TIA gr 3-5 1.0% 1.0% 1.0

Thromboembolic gr 3-5 0.8% 2.1% < 0.0001

Cardiac gr 3-5 2.1% 1.1% 0.0003

Ischemic heart disease gr 3-5 1.1% 0.6% 0.013

Cardiac failure gr 3-5 0.5% 0.1% 0.006

2626

Summary: Cardiovascular Events

• Compared with tamoxifen

- AIs reduce the risk of thromboembolic adverse events

- Adjuvant treatment with AIs has been associated with some increase in the

risk of CV events

• Current information is conflicting and insufficient to fully determine the

longer-term effect of AIs on CV health

• It is not possible at present to assign different cardiovascular risk profiles

to the individual AIs

• Further analyses of ongoing AI trials is required

2727

Bone Fractures

Letrozole Tamoxifen

Patients 3975 3988

Bone fractures 244 164

Patients w/ bone fracture 225 (5.7%) 159 (4.0%)

Bone fracture rate

(fracture/100 patient-years)

2.2 1.5

Risk ratio, p-value (L:T) 1.42 p=0.0006

2828

Endometrial Events

Letrozole Tamoxifen

Patients* 3089 3157

Endometrial biopsies (pts) 72 (2.3%) 288 (9.1%)

Invasive endometrial cancer 6 (0.2%) 15 (0.5%)

Invasive endometrial cancer

Risk ratio, p-value (L:T) 0.40, p=0.087

*Excludes 1717 patients with hysterectomy at baseline

2929

Perspective

Interpretation

Predictions

3030

RECURRENCE MORTALITY in trials of ~5 years of tamoxifen versus Not, ER+/unknown: 15-year outcome (life-table curve: 10386 women, all ages, 80% ER+, 30% N+)

Lancet May 14 2005

HRAbs RRn

HRAbs RRn

0.5710.4

0.6413.7

0.7411.8

0.703.6

0.707.9

0.749.2

EBCTCG 2000 (2005)Recurrence

EBCTCG 2000 (2005)Mortality

11.7 HR 0.44

14.8

3131

Smoothed hazard rates for recurrence? Start Early or Switch

0.5

1.0

1.5

2.0

2.5

3.0

0 1 2 3 4 5 6

Follow-up time (years)

Anastrozole Tamoxifen

0

Letrozole: Prevention of early distant relapses Should translate into mortality reduction

? Acquired Tamoxifen resistance developing at ~ 2-3 years

AnnualHR HR+

%

3232

0

0.5

1

1.5

2

2.5

3

1 2 3 4 5 6

Years since randomization

Ann

ual r

ates

%

AnastrozoleTamoxifen

0

Fracture rates over time

Fracture rates per 1000 women years: Anastrozole 22.6; Tamoxifen 15.6; P1 control 18.4; WHI control 19.1

3333

Cross trial Comparisons

Cross trial (indirect) comparisons may be unreliable:

- Different end-point definitions

- Different populations

- Different treatments

- Non randomised comparisons

They should be interpreted with caution both for efficacy and side effects comparisons

But they may lead to new hypotheses

3434

Endpoint: ComparisonsBIG 1-98 and ATAC

Favors TAM

Hazard ratio (LET:TAM)

DFS

OS

SDFS

Time to recurrence

DFS (w/o 2nd malignancy)

0.81

0.86

0.83

0.79

0.73 Time to distant metastasis (DDFS)

0.72

Favors LET

1.00.5 0.75 1.33 2.0

ATAC HR+68 mo1 33 mo2

1. Lancet. Jan 7, 2005.2. Lancet. June 22, 2002.

0.97

0.84

0.73

0.83 0.78

-

-

0.74

-

-

-

-

3535

Subgroups: OS

0.76

0.90

0.82

0.88

1.00

0.79

Favors L Favors T

1.00.5 0.75 1.33 2.0

Hazard Ratio (L:T)

CT given (n=2024)

CT not given (n=5986)

N-positive (n=3311)

N-negative (n=4174)

ER+ / PgR+ (n=5055)*

ER+ / PgR- (n=1631)*

* Based on local assessment

3636

Retrospective analysis of time to recurrence for ER/PgR subgroups

At risk:A 451 435 417 400 390 347 124T 429 412 375 353 327 276 96

Follow-up time (years)

25

0

5

10

15

20

0 1 2 3 4 5 6

Anastrozole (A)Tamoxifen (T)

Pat

ient

s (%

)

Patient group HR+ ER+PgR+ ER+PgR-

Hazard ratio 0.79 0.84 0.43

ER+/PgR-

3939

Protective Effect of Tamoxifenon cholesterol?

1) http://www.nhlbi.nih.gov/about/framingham/index.html; 2) McDonald CC et al.: BMJ 1995;311:977–80; 3) Rutqvist LE et al.:J Natl Cancer Inst 1993;85:1398–406, 4) Bradbury BD et al, Cancer March 2005, 5) Reis SE et al.; J Natl Cancer Inst 2001, Vol 93, No 1, Jan 3:16-21

Cardiovascular risk substantially and progressively

increases in women age >65 (Framingham study)1

The cardio protective effect of tamoxifen has been

studied in several trials

–The data are conflicting, some studies showed a

cardio protective effect2-4, others did not5.

4040

RECURRENCE MORTALITY in trials of ~5 years of tamoxifen versus Not, ER+/unknown: 15-year outcome (life-table curve: 10386 women, all ages, 80% ER+, 30% N+)

Lancet May 14 2005

HRAbs RRn

HRAbs RRn

0.5710.4

0.6413.7

0.7411.8

0.703.6

0.707.9

0.749.2

EBCTCG 2000 (2005)Recurrence

EBCTCG 2000 (2005)Mortality

11.7 HR 0.44

14.8

Questions and Answers