Trial Comparison:

1
Trial Comparison Arimidex, Tamoxifen,
Alone or in Combination (ATAC) and Breast
International Group (BIG) 1-98
2
Trial design and patient recruitment
3
Belgium 192 Czech Republic 84 France 366 Germany
121 Hungary 243 Ireland 41 Italy 654
Netherlands 195 Poland 107 Portugal 74 Slovakia
33 Spain 417 Sweden 291 Turkey 53 UK 3228
640
2222
160
30
201
14
9366 patients recruited from 381 centres in 21
countries
ATAC trial
4
(No Transcript)
5
ATAC trial design
9366 postmenopausal women with invasive breast
cancer mean age 64 years 84 hormone
receptor-positive 61 node negative 64 with
tumour ?2 cm in diameter
Surgery ? radiotherapy ? chemotherapy
Randomisation 111 for 5 years
Discontinued following initial analysis as no
efficacy or tolerability benefit compared with
tamoxifen arm
Anastrozole n3125
Tamoxifen n3116
Regular follow-up

• Secondary trial endpoints
• Incidence of contralateral breast cancer
• Time to distant recurrence
• Overall survival
• Time to breast cancer death

• Primary trial endpoints
• Disease-free survival
• Safety / tolerability

6
ATAC Completed Treatment Analysis

• Data cut-off 31 March 2004, based on at least 704
  deaths in the two monotherapy arms combined
• 68 months median follow-up beyond completion
  of treatment
• 59 months median treatment duration
• Only 8 of patients remain on treatment the
  great majority of these nearing completion

ATAC Trialists Group. Lancet 2005 365 60-62
7
BIG 1-98 trial design
R A N D O M I S E
Arm
Tamoxifen
A
8028 postmenopausal women with ER diseaseMedian
age 61 years52 node negative63 tumour ?2 cm
in diameter
B
Letrozole
Tamoxifen
Letrozole
C
Letrozole
Tamoxifen
D
0
2
5
1
3
4
Time (years)
A vs B March 1998 March 2000 (n1835) A vs B
vs C vs D September 1999 May 2003 (n 6193)
BIG Breast International GroupER estrogen
receptor-positive
Adapted from Thürlimann B. St Gallen
presentation 2005
8
ATAC 73 of patients have been followed-up for
5 years or more
Patients ()
Updated analysis (median follow-up 47 months)
Treatment completion analysis (median follow-up
68 months)
lt1
1lt2
2lt3
3lt4
4lt5
gt5
Duration of follow-up (years)
Total number of DFS events (monotherapy arms) 1226
DFS disease-free survival
9
BIG 1-98 only 15 of patients have been
followed-up for 5 years
Patients ()
?1
?2
?3
?4
?5
Follow-up (years)
Thürlimann B. St Gallen presentation 2005
10
Demographics
11
Patient characteristics
BIG 1-98 (n8010)
ATAC (n6291)
Age (years) Primary treatment () mastectomy ra
diotherapy chemotherapy
Mean 64.1 47.6 62.9 21.6
Median 61.0 43.0 71.6 25.3
ATAC Trialists Group. Lancet 2002 359
2131-39 Adapted from Thürlimann B. St Gallen
presentation 2005
12
Baseline disease characteristics
BIG 1-98 (n8010)
ATAC (n6291)
Tumour size ?2 cm () Nodal status
() node-positive node-negative unknown HR
status () ER/PgR ER/PgR- ER/PgR
unknown ER-/PgR
63.4 34.2 60.7 5.0 61.5 14.1 5.5 2.2
62.9 41.3 52.2 6.5 63.1 20.4 14.4 1.8
ATAC Trialists Group. Lancet 2002 359
2131-39 Adapted from Thürlimann B. St Gallen
presentation 2005
13
Efficacy analyses
14
Definition of disease-free survival differs

• ATAC
• loco-regional recurrence or new contralateral
  breast cancer (invasive or DCIS)
• distant recurrence or death (for any reason)
• BIG 1-98
• breast cancer recurrence (local, regional and
  distant) or invasive contralateral breast cancer
• non-breast cancer deaths (deaths without
  recurrence)
• non-breast cancer second primaries
• Time to recurrence is similar for both trials

ATAC Trialists Group. Lancet 2002 359 2131-39
Thürlimann B et al. The Breast 200514 S3.
Abstract S4
DCIS ductal carcinoma in situ
15
Definition of time to distantrecurrence appears
to differ

• ATAC - time to distant recurrence (TTDR)
• distant recurrence or any death following a
  loco-regional recurrence (including ipsilateral
  new breast cancer) or breast cancer death
• 45 of first events were distant events
• 18 of first events were locoregional
• BIG 1-98 - time to distant metastasis (TTM)
• breast cancer recurrence (excluding local or
  regional recurrences, and contralateral breast
  cancer)
• censoring for non-breast cancer deaths
• 65 of first events were distant events
• 12 of first events were local or regional

ATAC Trialists Group. Lancet 2005 365
60-62Thürlimann B et al. The Breast 200514 S3.
Abstract S4
excluding second primary events
16
ATAC disease-free survival (HR-positive
population)
25
Patients()
HR 0.83
95 CI (0.73, 0.94)
p value 0.005
A vs T
20
15
10
5
1.6
2.6
2.5
3.3
Absolute difference
0
0
1
2
3
4
5
6
Follow-up time (years)
At risk
A
2618
2540
2448
2355
2268
2014
830
T
2598
2516
2398
2304
2189
1932
774
Howell A. SABCS presentation 2004
CI, confidence interval
17
BIG 1-98 disease-free survival
97.7
95.1
90.5
86.8
84.0
Yearly
L
97.6
93.4
89.0
84.6
81.4
DFS
T
p value 0.003
HR 0.81
95 CI (0.70, 0.93)
L vs T
25
N8010
20
15
10
Letrozole Tamoxifen
5
0
0
1
2
3
4
5
Follow-up time (years)
At risk
4003
3892
2964
1261
892
567
L
4007
3896
2926
1238
866
544
T
Adapted from Thürlimann B. St Gallen
presentation 2005
18
ATAC recurrence (HR-positive population)
25
Patients()
p value 0.0002
HR 0.74
95 CI (0.64, 0.87)
A vs T
20
15
10
5
1.7
2.4
2.8
3.7
Absolute difference
0
0
1
2
3
4
5
6
Follow-up time (years)
At risk
A
2618
2540
2448
2355
2268
2014
830
T
2598
2516
2398
2304
2189
1932
774
ATAC Trialists Group. Lancet 200536560-62
19
BIG 1-98 breast cancer relapse(Time to
recurrence) Cumulative incidence
20
Proportion failure ()
5-year difference (L-T) -3.4?1.2 p0.0002
(based on CI)
13.6
15
Letrozole (L) Tamoxifen (T)
8.1
10
10.2
5
6.2
0
0
1
2
3
4
5
Years from randomisation
Thürlimann B. St Gallen presentation 2005
20
ATAC time to distant recurrence (HR-positive
population)
25
Patients()
p value 0.06
HR 0.84
95 CI (0.70, 1.00)
A vs T
20
15
10
5
0
0
1
2
3
4
5
6
Follow-up time (years)
At risk
A
2618
2550
2464
2386
2309
2051
845
T
2598
2533
2438
2361
2251
2005
816
Howell A. SABCS presentation 2004
21
ATAC overall survival (HR-positive population)
25
Patients()
p value 0.7
HR 0.97
95 CI (0.83, 1.14)
A vs T
20
15
10
5
0
0
1
2
3
4
5
6
Follow-up time (years)
At risk
A
2618
2566
2505
2437
2377
2117
867
T
2598
2549
2502
2430
2333
2080
855
Howell A. SABCS presentation 2004
22
ATAC efficacy summary(HR-positive population)
Hazardratio
Disease-free survival
0.83
Time to recurrence
0.74
Time to distant recurrence
0.84
Overall survival
0.97
Time to breast cancer death
0.87
Contralateral breast cancer
0.47
0.2
0.4
0.6
0.8
1.0
1.2
1.5
2.0
Hazard ratio (AT) and 95 CI
Anastrozole (A) better
Tamoxifen (T) better
ATAC Trialists Group. Lancet 200536560-62
23
ATAC efficacy analysis (ITT and HR ve)
ITT
HR
0.83 0.74 0.84 0.97 0.87 0.47
0.87 0.79 0.86 0.97 0.88 0.58
Disease-free survival
Time to recurrence
Time to distant recurrence
Overall survival
Time to breast cancer death
Contralateral breast cancer
0.2
0.4
0.6
0.8
1.0
1.2
1.5
2.0
HR (AT) and 95 CI
ITT population
Anastrozole (A) better
Tamoxifen (T) better
HR ve population
ATAC Trialists Group. Lancet 200536560-62
24
BIG 1-98 efficacy summary
Hazard ratio
BIG 1-98
ATAC
Disease-free survival
0.81
Time to recurrence
0.72
0.74
Time to distant recurrence
0.73
0.84
Overall survival
0.86
0.97
Systemic disease-free survival
0.83
Disease-free survival (without 2nd primary)
0.79
0.83
0.2
0.4
0.6
0.8
1.0
1.2
1.5
2.0
Hazard ratio (LT) and 95 CI
Letrozole (L) better
Tamoxifen (T) better
Adapted from Thürlimann B. St Gallen
presentation 2005
25
BIG 1-98 sites of first failure
Letrozole ()
Tamoxifen ()
p value
Failures (DFS events) local contralateral
breast regional distant second (non-breast)
malignancy death without recurrence Deaths System
ic failures
8.8 0.5 0.4 0.3 4.4 1.7 1.4 4.1 8.1
10.7 0.9 0.7 0.3 5.8 2.0 0.9 4.8 9.6
0.004 0.047 0.125 0.845 0.006 0.324 0.077 0.176 0.
020
Regional includes axilla or internal
mammarySDFS ignores local and contralateral
events
Thürlimann B. St Gallen presentation 2005
26
ATAC vs BIG 1-98 efficacy summary

• Anastrozole is more effective than tamoxifen in
  reducing the risk of recurrence, distant
  recurrence and contralateral breast cancer
• absolute difference between anastrozole and
  tamoxifen continues to increase over time, and
  extends beyond completion of treatment
• Letrozole demonstrates DFS benefits and early
  benefits in distant recurrence
• BIG 1-98 has a comparatively higher number of
  patients per arm, resulting in a higher number of
  events per unit time
• patient population in BIG 1-98 has a slightly
  worse prognosis
• Absolute differences at 5 years for BIG 1-98 data
  are projected out to 5 years and are calculated
  from immature data hence liable to change

27
Sub-group analysis
28
ATAC time-to-recurrence by subgroup
All patients
0.40
1.50
1.75
0.60
0.80
1.00
1.25
Intent-to-treat population
Hazard ratio (AT) and 95 CI
Anastrozole (A) better
Tamoxifen (T) better
Howell A. SABCS presentation 2004
29
BIG 1-98 disease-free survivalby subgroup
All patients
0.40
1.50
1.75
0.60
0.80
1.00
1.25
Intent-to-treat population
Hazard ratio (LT) and 95 CI
Letrozole (L) better
Tamoxifen (T) better
Adapted from Thürlimann B. St Gallen
presentation 2005
30
ATAC vs BIG 1-98 subgroup summary (1)

• Anastrozole demonstrated advantages over
  tamoxifen for all subgroups examined
• no heterogeneity of subgroups
• no significant interaction with any baseline
  prognostic factor, including prior chemotherapy
  or nodal status
• more effective than tamoxifen in overall HRve
  group
• even greater improvement in ERPgR- subgroup
• Subgroup analyses must be interpreted with
  caution
• should not be used as a basis for making clinical
  decisions

31
ATAC vs BIG 1-98 subgroup summary (2)

• Letrozole demonstrated benefits over tamoxifen
• node positive patients
• no apparent benefit in node negative patients
• prior chemotherapy patients
• slightly worse prognosis, more patients received
  prior chemotherapy (25 vs 20)
• No apparent difference between ER/PgR and
  ER/PgR- subgroups for letrozole and tamoxifen
• tamoxifen does not appear to be performing in
  line with expectations
• previous studies demonstrate that ER/PgR-
  patients on tamoxifen have a higher rate of
  recurrence than ER/PgR
• Subgroup analyses must be interpreted with
  caution
• should not be used as a basis for making clinical
  decisions

32
Tolerability analysis
33
BIG 1-98 safety analysis

• Included all patients that had received at least
  1 treatment dose
• Protocol-specified only targeted adverse event
  data was collected every 6 months
• Number of patients experiencing at least 1
  serious adverse event
• 587 vs 643 (letrozole vs tamoxifen)

34
ATAC overview of adverse events
All adverse events Adverse events leading to
withdrawal Drug-related adverse events leading
to withdrawal All serious adverse events Serious
adverse events leading to withdrawal Serious
adverse events leading to death Drug-related
serious adverse events leading to death
p value 0.2 0.0002 0.0005 0.03 0.04 0.6 0.5
Tamoxifen ()(n3094) 94.6 14.3 8.9 36.0
5.9 3.6 0.3
Anastrozole ()(n3092) 93.9 11.1 6.5 33.3
4.7 3.3 0.2
Adverse events on treatment or within 14 days of
discontinuation
Howell A. SABCS presentation 2004
35
ATAC pre-defined adverse events
Completion analysis ()
p value
T 40.9 10.2 13.2 0.8 2.8 4.5 2.4 29.4 7.7
A 35.7 5.4 3.5 0.2 2.0 2.8 1.6 35.6 11.0
lt0.0001 lt0.0001 lt0.0001 0.02 0.03 0.0004 0.02 lt
0.0001 lt0.0001
Hot flushes Vaginal bleeding Vaginal
discharge Endometrial cancer Ischaemic
cerebrovascularevent Venous thromboembolicevents
Deep venous thromboembolic events Joint
symptoms Total fractures
Adverse events on treatment or within 14 days of
discontinuation Excludes patients with prior
hysterectomy and includes on- and off-therapy
AEs Fractures occurring at anytime prior to
recurrence (includes patients no longer receiving
treatment)
ATAC Trialists Group. Lancet 200536560-62
36
BIG 1-98 targeted adverse events
Primary core analysis ()
T 38.1 6.6 16.2 9.5 2.4 2.6 1.1 8.3 4.1 19.2
L 33.6 3.3 14 8.8 1.0 2.7 1.2 8.7 5.8 43.6
Hot flushes Vaginal bleeding Night sweats Nausea
Thromboembolic events Vomiting CVA/TIA Other
cardiovascular Bone fracture Hypercholesterolemia
No arthralgia/joint symptoms or osteoporosis data
are available from BIG 1-98 Endometrial cancer
shows no significant difference between L and T
Adapted from Thürlimann B. St Gallen
presentation 2005
37
ATAC vs BIG 1-98 bone fractures
ATAC (A vs T)
BIG 1-98 (L vs T )
Patients
Patients with bone fracture
340 vs 237 (11.0 vs 7.7) 1.49, plt0.0001
228 vs 162 (5.8 vs 4.1) 1.44, p0.0006
Odds ratio, p value
2.2 vs 1.5 (per 100 patient years)
2.3 vs 1.6 (per 100 patient years)
Bone fracture rate
Adapted from ATAC Trialists Group. Lancet
200536560-62 Thürlimann B. St Gallen
presentation 2005
38
ATAC fracture risk is predictable and manageable
3
Annual rates,
2.5
2
1.5
1
Anastrozole 1 mg od Tamoxifen 20 mg od
0.5
0
1
2
3
4
5
6
0
Years since randomisation
Number at risk
Years Arimidex Tamoxifen
0 3092 3094
1 2923 2932
2 2724 2741
3 2553 2579
4 2393 2401
5 2070 2100
6 845 846
Calculated using Kaplan-Meier estimates
Howell A. SABCS presentation 2004
39
ATAC vs BIG 1-98 endometrial cancer
ATAC (A vs T)
BIG 1-98 (L vs T )
Patients
Patients with endometrial cancer
5 vs 17 (0.2 vs 0.8) 0.29, p0.02
6 vs 15 (0.2 vs 0.4) 0.40, p0.078
Odds ratio, p value
Adapted from ATAC Trialists Group. Lancet
200536560-62 Thürlimann B. St Gallen
presentation 2005
40
BIG 1-98 Grade 3-5 cardiovascular events
Letrozole(n3965)
Tamoxifen(n3984)
Patients
CVA/TIA Thromboembolic Other cardiovascular
46 (1.2) 30 (0.8) 143 (3.6)
42 (1.1) 79 (2.0) 101 (2.5)
There is a significantly higher number of other
cardiovascular events on letrozole compared with
tamoxifen (p0.006)
Thürlimann B. St Gallen presentation 2005
41
BIG 1-98 death without recurrence Cumulative
incidence
20
Proportion failure ()
5-year difference (L-T) 1.3?0.6 p0.08 (based
on CI)
15
Letrozole (L) Tamoxifen (T)
10
3.1
5
1.4
1.8
0
0.8
1
2
3
4
5
0
Years from randomisation
Thürlimann B. St Gallen presentation 2005
42
BIG 1-98 deaths without recurrence(non-breast
cancer deaths)
Letrozole(n4003)
Tamoxifen(n4007)
Patients
Total CVA thromboembolic cardiac other Overall
p value based on cumulative incidence
55 7 3 26 19
38 1 2 13 22
0.08
In ATAC, the numbers of cardiovascular deaths are
comparable between anastrozole and tamoxifen (49
vs. 46, respectively)
Adapted from Thürlimann B. St Gallen
presentation 2005
43
ATAC deathsMedian follow-up 68 months
Anastrozole(n3125)
Tamoxifen(n3116)
Patients
All deaths non-breast cancer deaths cerebrovascu
lar cardiac
411 176 14 49
420 155 21 46
A detailed review found that the non-breast
cancer deaths in the anastrozole arm were due to
a variety of apparently unrelated causes, with
no link to anastrozole
ATAC Trialists Group. Lancet 2005
36560-62 ATAC Trialists Group. Lancet 2005 In
Press
44
Comparison of safety between ATAC and BIG 1-98
Compared with tamoxifen
L NS ? ? ? ? ? ? ? ? ?

• A
• ?
• ?
• ?
• NS
• ?
• ?
• ?
• ?
• ? ?

Endometrial cancer Risk of stroke Venous
thromboembolic events Cardiovascular deaths Joint
symptoms Fractures Hot flushes Vaginal
bleeding Vaginal discharge Hysterectomy
? not reported
45
ATAC tolerability and safety summary vs tamoxifen

• Compared with tamoxifen, anastrozole is
  associated with significantly fewer
• SAEs, treatment-related AEs and withdrawals due
  to SAEs or AEs
• potentially life-threatening AEs such as
  endometrial cancer, thromboembolic and
  cerebrovascular events
• No new safety concerns have emerged with
  long-term follow-up. There is no issue with
  cardiovascular safety
• Anastrozole now has a known, predictable and
  manageable safety profile

Only anastrozole has a tolerability profile of
this robustness and maturity, as it covers more
than 5-years follow-up
AEs adverse eventsSAEs serious AEs
The ATAC Trialists Group. Lancet 2005 365
60-62 Howell A. Breast Cancer Res Treat 2004
88 (Suppl 1) S7, abs 1
46
BIG 1-98 tolerability and safety summary vs
tamoxifen

• Serious safety concerns about letrozole have
  emerged in this first analysis
• increased incidence of stroke and cardiovascular
  events
• increase in number of cerebrovascular and
  cardiovascular deaths
• No significant reduction in the incidence of
  endometrial cancer was observed
• The long-term safety profile of letrozole is
  unknown at this stage
• cardiovascular effects of letrozole require
  further evaluation

BIG 1-98 has raised serious safety concerns for
letrozoleat this early stage
47
Summary
48
Conclusions (1)

• The ATAC completed treatment analysis extends and
  strengthens the evidence that 5 years of
  anastrozole is significantly more effective and
  better tolerated than 5 years of tamoxifen
• The efficacy benefit continues to increase with
  time and extends beyond the completion of therapy
• These data support using anastrozole as initial
  adjuvant therapy
• The higher rates of recurrence, adverse events,
  and withdrawals from treatment with tamoxifen and
  the substantial benefit of anastrozole over the
  first3 years justify the approach of offering
  the most effective therapy at the earliest
  opportunity

49
Conclusions (2)

• BIG 1-98 provides further evidence that tamoxifen
  should no longer be the standard of care for EBC
• No overall efficacy benefits have emerged for
  letrozole in BIG 1-98 that have not already been
  demonstrated for anastrozole in the ATAC trial
• There appear to be marked differences emerging in
  the safety of the aromatase inhibitors in the
  adjuvant setting
• women treated with letrozole have a greater risk
  of stroke and cardiac events
• The ATAC completed treatment analysis
  demonstrates that the overall benefitrisk
  profile remains clearly and consistently in
  favour of anastrozole
• Only anastrozole has established efficacy and
  safety withgt5 years long-term follow-up data

50
Back-up slides
51
ATAC patient characteristics
Tamoxifen (n3116)
Anastrozole (n3125)
Mean age (years) Receptor status
() positive negative unknown Primary
treatment () mastectomy radiotherapy chemother
apy
64.1 83.7 8.3 8.0 47.8 63.3 22.3
64.1 83.4 8.7 7.9 47.3 62.5 20.8
ATAC Trialists Group. Lancet 2002 359 2131-39
52
BIG 1-98 patient characteristics
Tamoxifen (n4007)
Letrozole (n4003)
Median age (years) Primary treatment
() chemotherapy Surgery/RT group () BC with
RT BC without RT mastectomy with RT mastectomy
without RT
61.0 25.3 53.3 2.8 18.3 25.4
61.0 25.3 54.0 3.3 17.6 24.8
BC breast conservation RT radiotherapy
Adapted from Thürlimann B. St Gallen
presentation 2005
53
ATAC baseline disease characteristics
Tamoxifen (n3116)
Anastrozole (n3125)
Primary tumour size () T1 (?2 cm) Nodal status
() node-positive node-negative node-unknown HR
status () ER/PgR ER/PgR- ER/PgR
unknown ER-/PgR
63.9 34.9 60.0 5.0 61.8 14.4 5.3 2.0
62.9 33.6 61.5 4.9 61.1 13.8 5.8 2.4
HR hormone receptor ER oestrogen receptor
PgR progesterone receptor
ATAC Trialists Group. Lancet 2002 359 2131-39
54
BIG 1-98 baseline disease characteristics
Tamoxifen (n4007)
Letrozole (n4003)
Tumour size ?2 cm () Nodal status
() node-positive node-negative unknown HR
status () ER/PgR ER/PgR- ER/PgR
unknown ER-/PgR
63.5 41.5 52.0 6.5 63.5 20.2 14.5 1.5
62.3 41.2 52.3 6.5 62.7 20.5 14.3 2.1
Adapted from Thürlimann B. St Gallen
presentation 2005
55
ATAC vs BIG 1-98 demographics

• Patients in the ATAC trial had an improved
  prognosis compared with patients in BIG 1-98
• fewer patients in ATAC had node positive disease
• fewer patients in ATAC had received prior
  radiotherapy
• fewer patients in ATAC had received prior
  chemotherapy

56
Definition of further ATAC endpoints

• Time to recurrence (TTR)
• loco-regional recurrence (including ipsilateral
  new breast cancer) or new contralateral breast
  cancer
• distant recurrence or death due to breast cancer
• Overall survival (OS)
• death (for any reason)
• Time to breast cancer death (TTBCD)
• any death following a loco-regional (including
  ipsilateral new breast cancer) or distant
  recurrence
• breast cancer death

57
Definition of further BIG 1-98 endpoints

• Time to recurrence (TTR)
• breast cancer recurrence or new contralateral
  breast cancer (excluding non-breast cancer second
  primaries)
• censoring for non-breast cancer deaths
• Overall survival (OS)
• death (for any reason)
• DFS without second primary events
• as DFS (excluding non-breast second primaries)
• Systemic disease-free survival (SDFS)
• regional or distant recurrence (not including
  local and contralateral)
• non-breast second primaries
• non-breast cancer death

allows comparison with ATAC no ATAC equivalent
Thürlimann B et al. The Breast 200514 S3.
Abstract S4
58
ATAC efficacy analysis (ITT and HR ve)
ITT
HR
0.83 0.74 0.84 0.97 0.87 0.47
0.87 0.79 0.86 0.97 0.88 0.58
Disease-free survival
Time to recurrence
Time to distant recurrence
Overall survival
Time to breast cancer death
Contralateral breast cancer
0.2
0.4
0.6
0.8
1.0
1.2
1.5
2.0
HR (AT) and 95 CI
ITT population
Anastrozole (A) better
Tamoxifen (T) better
HR ve population
ATAC Trialists Group. Lancet 200536560-62
59
ATAC recurrence in ER/PgR- patients
25
Patients ()
Anastrozole (A)
Tamoxifen (T)
20
15
10
5
0
0
1
2
3
4
5
6
Follow-up time (years)
At risk
A
451
435
417
400
390
347
124
T
429
412
375
353
327
276
96
60
ATAC recurrence in ER/PgR patients
Patients ()
At risk
A
1930
1880
1820
1755
1690
1505
641
T
1904
1849
1779
1716
1639
1459
597
61
ATAC retrospective analysis ofHR subgroups

• Anastrozole was more effective than tamoxifen in
  the overall HR group
• The improvement with anastrozole in the ERPgR
  subgroup was comparable to that for the HR group
• There was even greater improvement with
  anastrozole in the ERPgR- subgroup
• The relative benefit for anastrozole over
  tamoxifen appears to be larger in patients with
  ERPgR- tumours than in those with ERPgR
  tumours, but prospective studies are needed to
  confirm this

Patient group HR ER/PgR
ER/PgR- Hazard ratio 0.79
0.84 0.43
62
BIG 1-98 DFS in ER/PgR subgroups
ER PgR (n5055)
ER PgR- (n1631)
ER PgR unknown (n1154)
0.5
0.75
1.0
1.33
2.0
Hazard Ratio (AT) and 95 CI
Letrozole better
Tamoxifen (T) better
Based on local assessment
Thürlimann B. St Gallen presentation 2005
63
Introduction to best first

• The risk of recurrence is highest in the first
  five years after surgery, with a peak at 2 years
• Patients deserve to receive the best treatment
  first in order to reduce the risk of breast
  cancer recurrence

64
When to treat?

• Recurrence rates in early breast cancer are
  highest in the first 5 years after surgery, with
  a peak at 2 years, regardless of baseline
  prognostic factors
• Tamoxifen is associated with higher rates of
  recurrence, AEs and withdrawals than anastrozole
• Substantial benefit with anastrozole in the first
  3 years justifies offering the most effective
  therapy at the earliest opportunity

65
Most recurrences occur within the first 5 years
of primary therapy
Recurrence rate/year()
Need to give most effective treatment firstto
reduce risk of recurrence
Year
Saphner et al JCO 1996 14 2738-2746
66
Annual risk of recurrence ECOG data
Hazard of recurrence by yearly interval ()
Time (years)
ECOG Eastern Cooperative Oncology Group
Saphner T et al. J Clin Oncol 1996142738-2746
67
Timing of recurrence in the first10 years
post-diagnosisAdapted from EBCTCG metaanalysis
80
Proportion of recurrence ()
Node ve
62
61
Node -ve
60
38
39
40
20
0
0-5 years
5-10 years
EBCTCG, Lancet 1998 351 1451-1467
68
ATAC smoothed hazard rates for recurrence
(HR-positive population)
3.0
Annualhazardrates()
2.5
2.0
1.5
1.0
Anastrozole
Tamoxifen
0.5
0
0
1
2
3
4
5
6
Follow-up time (years)
69
Patients deserve the best treatment first

• Recurrence rates in early breast cancer are
  highest in the first 5 years after surgery, with
  a peak at2 years, for all patients (irrespective
  of risk)
• Anastrozole demonstrated substantial benefits
  over tamoxifen throughout the entire 5-year
  follow-up period in the ATAC trial, regardless of
  baseline prognostic factors
• the peak of recurrences at years 1-3 is
  suppressed by anastrozole
• All patients deserve the best treatment available
  at the earliest opportunity in order to reduce
  the risk of recurrence