| Three-fourths of breast cancers ER positive and account for most breast cancer deaths |
Not including bone measurements (DXA scan) as predetermined endpoints, instead relying on self-report |
|
| Strong rationale based on contralateral breast cancer reduction in adjuvant trials |
|
|
| Placebo-controlled design |
Lack of active comparator (e.g., raloxifene) to determine best hormonal strategy: no estrogen at all vs. best balance between agonistic and antagonistic effects |
|
| Eligibility used a pragmatic and clinically relevant algorithm: |
Loose definition of “high risk,” especially including all women ≥60 years of age, even those with Gail model risk ≤1.66% |
| 1. age ≥60 y |
| 2. Gail model |
| 3. history of IEN |
|
| Fast recruitment |
Study immaturity: powered for 38 invasive cancers for final analysis with 3-year recruitment +1.2-y follow-up. Although technically acceptable, probably not clinically meaningful. Follow-up too short for safety and risk:benefit assessment |
|
| Main findings: predictable high activity and excellent safety/tolerability profile |
Crossover to exemestane is being offered to women on placebo. Resulting contamination diminishes the value of follow-up, making MAP.3 a large proof-of-principle trial (activity vs. efficacy). |
