Arimidex Tamoxifen Alone or in Combination ATAC trial

‘Arimidex’, Tamoxifen, Alone or in Combination (ATAC) trial: Completed Treatment Analysis

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 â Surgery radiotherapy chemotherapy â Randomization 1: 1: 1 for 5 years å ‘Arimidex’ n=3125 â Tamoxifen n=3116 â æ Discontinued following initial Combination analysis as no efficacy or n=3125 tolerability benefit compared with tamoxifen arm Regular follow-up â Primary trial endpoints: • Disease-free survival • Safety / tolerability Secondary trial endpoints: • Incidence of contralateral breast cancer • Time to distant recurrence • Overall survival • Time to breast cancer death

ATAC trial analysis history First analysis – Dec 2001 Median follow-up : 33 months 1 Updated analysis – November 2002 Median follow-up : 47 months 2 Treatment Completion analysis – November 2004 Median follow-up : 68 months Women years’ follow up: 49, 941 Total events: 1867 1. The ATAC Trialists’ Group. Lancet 2002; 359: 2131– 2139 2. The ATAC Trialists’ Group. Cancer 2003; 98: 1802 – 1810 3. The ATAC Trialists’ Group. Lancet 2005; 365: 60 -62

ATAC completed treatment analysis l Data cut off: 31 st March 2004: – prospectively defined based on at least 704 deaths in the two monotherapy arms combined l Follow-up: – 68 months’ median follow-up – i. e. extends beyond completion of treatment – 59 months’ median treatment duration – Only 8% of patients remain on treatment – the great majority of these nearing completion

Efficacy analysis

Disease-free survival Curves shown for HR+ patients 17% RR HR# 95% CI p-value HR+* 0. 83 (0. 73– 0. 94) 0. 005 ITT** 0. 87 (0. 78 -0. 97) 0. 01 25 Patients (%) 20 15 ‘Arimidex’ (A) Tamoxifen (T) 10 5 Absolute difference: 0 0 At risk: A 2618 T 2598 1 2 1. 6% 2. 6% 3 4 2. 5% 3. 3% 5 6 2014 1932 830 774 Follow-up time (years) 2540 2516 2448 2398 2355 2304 2268 2189 *ITT = Intention-to-treat population; **HR+ = Hormone receptor-positive patients; #A vs T

Time to Recurrence Curves shown for HR+ patients 25 HR (0. 64– 0. 87) 0. 0002 ITT Patients (%) 0. 74 0. 79 (0. 79 -0. 90) 26% RR p-value HR+ 20 95% CI 0. 0005 15 ‘Arimidex’ (A) Tamoxifen (T) 10 5 Absolute difference: 0 0 1 2 1. 7% 2. 4% 3 2. 8% 4 3. 7% 5 6 Follow-up time (years) At risk: A 2618 2540 2448 2355 2268 2014 830 T 2598 2516 2398 2304 2189 1932 774

Smoothed hazard rates for recurrence (HR+ population) Annual 3. 0 hazard rates 2. 5 (%) 2. 0 1. 5 1. 0 ‘Arimidex’ (A) Tamoxifen (T) 0. 5 0 0 1 2 3 4 Follow-up time (years) 5 6

Time to Distant Recurrence Curves shown for HR+ patients 25 HR Patients (%) 0. 84 (0. 70– 1. 00) 0. 06 ITT 15 p-value HR+ 20 95% CI 0. 87 (0. 74 -0. 99) 0. 04 ‘Arimidex’ (A) Tamoxifen (T) 10 5 0 0 1 2 3 4 5 6 A 2618 2550 2464 2386 2309 2051 845 T 2598 2533 2438 2361 2257 2005 816 At risk:

Overall Survival* Curves shown for HR+ patients HR 95% CI p-value HR+ 0. 97 (0. 83– 1. 14) 0. 7 ITT 0. 97 (0. 85 -1. 12) 0. 7 25 Patients (%) 20 15 ‘Arimidex’ (A) Tamoxifen (T) 10 5 0 0 At risk: A 2618 T 2598 1 2 3 4 5 6 Follow-up time (years) 2566 2505 2437 2377 2117 867 2549 2502 2430 2333 2080 855 * Includes non breast cancer deaths

Time to Breast Cancer Death Curves shown for HR+ patients 25 HR Patients (%) p-value HR+ 0. 87 (0. 70– 1. 09) 0. 2 ITT 20 95% CI 0. 88 (0. 74 -1. 05) 0. 2 15 ‘Arimidex’ (A) Tamoxifen (T) 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

Incidence of new (contralateral) breast primaries in HR+ population HR Number 60 of cases 50 p-value 0. 47 A vs T 95% CI (0. 29– 0. 75) 0. 001 54 40 30 26 20 10 0 ‘Arimidex’ (A) (n=2618) Tamoxifen (T) (n=2598)

Efficacy analysis – ITT and HR+ populations ITT population HR+ population Disease-free survival Time to recurrence Time to distant recurrence Overall survival Time to breast cancer death Contralateral breast cancer 0. 2 *ITT= intent-to-treat *HR+=hormone receptor-positive 0. 4 0. 6 0. 8 1. 0 1. 2 1. 5 2. 0 Hazard Ratio (A: T) and 95% CI ‘Arimidex’ (A) better Tamoxifen (T) better

Analysis of time to recurrence for subgroups of the ITT* population Nodal status +ve -ve unknown Tumour size ≤ 2 cm >2 cm unknown* Receptor status +ve -ve unknown Previous chemotherapy yes no All patients Hazard ratio (A: T) and 95% CI 0. 40 0. 60 0. 80 ‘Arimidex’ better *Confidence limit extends beyond plot 1. 00 1. 25 1. 50 1. 75 Tamoxifen better

Summary of efficacy endpoints l In the HR+ population, compared with tamoxifen, ‘Arimidex’ reduces the risk of : – all events: 17% (p=0. 005) – recurrence: 26% (p=0. 0002) – distant recurrence: 16% (p=0. 06) – contralateral recurrence: 53% (p=0. 001) l There is also a (non-significant trend for a reduction in breast cancer mortality: 13% (p=0. 2)

‘Arimidex’ demonstrates superior efficacy to tamoxifen l ‘Arimidex’ is more effective than tamoxifen in reducing the risk of recurrence, distant recurrence and contralateral breast cancer l The absolute difference between ‘Arimidex’ and tamoxifen continues to increase over time, and extends beyond completion of treatment l As expected, overall survival is similar for both treatments, i. e. ‘Arimidex’ maintains the significant survival benefit already established for tamoxifen, with a trend in favour of ‘Arimidex’ for breast cancer death l There are no significant subgroup interactions

Added benefit versus tamoxifen HR+ve population Reduction in risk of recurrence Reduction in risk of breast cancer mortality Reduction in risk of contralateral breast cancer EBCTCG ATAC Benefit for tamoxifen vs placebo Additional benefit of ‘Arimidex’ vs tamoxifen 50% 26% 28% 13% 47%* 53% *hormone receptor-positive and -negative patients EBCTCG = Early Breast Cancer Trialists’ Collaborative Group

Added benefit versus tamoxifen 38% risk of recurrence with no adjuvant treatment 50% risk reduction with tamoxifen Further 26% risk reduction with ‘Arimidex’

When to treat? l 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 l Tamoxifen is associated with higher rates of recurrence, AEs and withdrawals than ‘Arimidex’ l Substantial benefit with ‘Arimidex’ in the first 3 years justifies offering the most effective therapy at the earliest opportunity Best treatment first !

Annual hazard of recurrence peaks at 2 years regardless of baseline prognostic factors Need to give most effective treatment first to reduce risk of recurrence Adapted from Saphner et al, 1996

Tolerability analysis

Overview of adverse events* 'Arimidex' (%) (n=3092) Tamoxifen (%) (n=3094) p-value Adverse events leading to withdrawal 11. 1 14. 3 0. 0002 Drug-related adverse events leading to withdrawal 6. 5 8. 9 0. 0005 All serious adverse events 33. 3 36. 0 0. 03 Serious adverse events leading to withdrawal 4. 7 5. 9 0. 04 Serious adverse events leading to death 3. 3 3. 6 0. 6 Drug-related serious adverse events leading to death 0. 2 0. 3 0. 5 *Adverse events on treatment or within 14 days of discontinuation

Pre-defined adverse events* Completion analysis Hot flushes Vaginal bleeding Vaginal discharge Endometrial cancer** Ischaemic cerebrovascular event Venous thromboembolic events Deep venous thromboembolic events Joint symptoms Total fractures*** p-value A 35. 7 5. 4 3. 5 0. 2 2. 0 T 40. 9 10. 2 13. 2 0. 8 2. 8 <0. 0001 0. 02 0. 03 2. 8 4. 5 0. 0004 1. 6 2. 4 0. 02 35. 6 11. 0 29. 4 7. 7 <0. 0001 *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)

Tolerability summary l Compared with tamoxifen, 'Arimidex' is associated with significantly fewer: – SAEs, treatment-related AEs and withdrawals – potentially life threatening AEs such as endometrial cancer, thromboembolic, and cerebrovasular events l No new safety concerns with long-term follow-up l Only 'Arimidex' has a tolerability profile this robust and mature, covering the full 5 year treatment period l 'Arimidex' now has a known, predictable and manageable safety profile Tolerability profile not only different but SUPERIOR

Summary

ATAC summary l ATAC Completed Treatment Analysis extends and strengthens the evidence that 5 years of 'Arimidex' is significantly more effective and better tolerated than 5 years of tamoxifen l Overall risk: benefit profile remains clearly in favour of 'Arimidex' l The absolute benefits for 'Arimidex' over and above those of tamoxifen continue to increase with time and extend beyond the completion of therapy

ATAC in context l 'Arimidex' is a more effective and better-tolerated adjuvant treatment than tamoxifen l These findings provide a basis for establishing 'Arimidex' as the standard of care for the initial 5 years’ adjuvant treatment of postmenopausal women with hormone receptor-sensitive early breast cancer Howell, SABCS 2004

Big News ! Daily Mail Daily Mirror The Times UK Headlines 08 Dec 2004 Independent Evening Standard