NSABP B-35

Background

Phase III, double-blind, placebo-controlled RCT. 3104 postmenopausal women with hormone-receptor-positive DCIS after BCS with clear margins, all receiving whole-breast radiation. Randomized to anastrozole vs tamoxifen for 5 years — the first head-to-head comparison of an AI vs tamoxifen in the DCIS setting. Extends the AI benefit seen in invasive BC prevention (IBIS-II) and adjuvant therapy to the DCIS population.

Interventions and follow up

Arm A: Anastrozole 1mg daily × 5 years
Arm B: Tamoxifen 20mg daily × 5 years
Primary endpoint: Breast cancer-free interval (BCFI)
mFollow up: 9 years

Results

BCFI (10-yr): 93.5% vs 89.2% (anastrozole vs tamoxifen), HR 0.73, 95% CI 0.56–0.96, P=.0234
Age <60 subgroup: benefit concentrated here; HR favored anastrozole
Age ≥60 subgroup: no significant BCFI difference between arms
Invasive BC events: HR 0.68 (anastrozole vs tamoxifen)
Contralateral BC: HR 0.52 (favoring anastrozole)

Adverse events

Musculoskeletal: arthralgia/musculoskeletal symptoms higher with anastrozole (53.1% vs 35.4%); bone fractures anastrozole 8.5% vs tamoxifen 7.1% (not significant)
Vasomotor: hot flashes similar between arms
Gynecologic/thromboembolic: endometrial/uterine cancer lower with anastrozole (0 vs 0.8%, P=.004); DVT/PE lower with anastrozole

Conclusions

Anastrozole significantly improved breast cancer-free interval compared to tamoxifen in postmenopausal women with hormone-receptor-positive DCIS after BCS, with the greatest benefit in women under 60. Anastrozole also reduced uterine cancer and VTE risk vs tamoxifen, suggesting a favorable benefit-risk shift for AIs in DCIS.

Key Limitations

Benefit concentrated in women <60 — postmenopausal women ≥60 did not show a statistically significant BCFI improvement with anastrozole. Primary endpoint is BCFI (composite) — the trial was not individually powered for OS or invasive BC mortality. Musculoskeletal toxicity (53% with anastrozole) significantly exceeds tamoxifen and may drive non-adherence in practice. Generalizability to higher-grade DCIS or positive margins is uncertain. All patients received radiation (RTOG 9804) — the relative contribution of endocrine therapy vs RT is not disentangled.

Clinical Context

NSABP B-35 supports anastrozole as a preferred endocrine option for hormone-receptor-positive DCIS in postmenopausal women, particularly those under 60. ASCO endorses either an AI (anastrozole or exemestane) or tamoxifen as adjuvant endocrine therapy in DCIS; anastrozole avoids uterine cancer and VTE risk, making it preferable for many postmenopausal women. For premenopausal DCIS, tamoxifen remains the only endocrine option. ESMO-MCBS: not applicable (DCIS, local-recurrence endpoint).

References