Menopause HRT and breast cancer: what the two WHI trials actually found

Menopausal HRT does not have one transferable breast-cancer risk number. In WHI, combined oral conjugated equine estrogens plus medroxyprogesterone acetate (CEE+MPA) increased breast-cancer incidence versus placebo among women with a uterus, while oral CEE alone was associated with lower incidence versus placebo among women who had undergone hysterectomy. These were separate randomized populations, not a head-to-head choice between regimens. FDA’s 2026 boxed-warning edits did not remove product-specific breast-cancer contraindications or warnings.¹²³

The first useful question is therefore not “Is estrogen safe?” It is which product, combination, route, population, timing and outcome a claim describes. A practice that cites a reassuring estrogen-alone result for someone considering estrogen plus a progestogen has crossed a trial boundary.

Keep the two WHI populations in separate columns

The original long-term analysis reported these results with confidence intervals of 1.13–1.45 for the combined-therapy hazard ratio and 0.65–0.93 for the estrogen-alone ratio. They are group-level incidence estimates for the studied oral regimens and follow-up, not individual probabilities or a calculation for a different modern product.¹

The table is deliberately not a menu from which a person can select CEE alone because its column looks lower. Uterus status determined which WHI trial a woman could enter, and unopposed systemic estrogen has a separate endometrial-cancer issue in women with a uterus. Trial design, product, dose and clinical history belong in a prescriber’s review.

What the WHI comparison cannot answer

WHI did not directly randomize a woman to contemporary transdermal estradiol plus micronized progesterone versus another modern combination in an otherwise identical population. Its results cannot prove that every estrogen route or progestogen shares the same breast-cancer effect; they also cannot prove a formulation is risk-free because it differs from CEE+MPA.

Observational studies may explore product and route differences, but they have different confounding and cannot silently replace the randomized result. When a clinic uses the phrase “bioidentical means no breast risk,” ask for the exact finished product and a study in a comparable population with breast-cancer outcomes. A chemical descriptor does not supply that evidence.

What changed in the FDA box—and what remains

In February 2026 FDA approved specified label changes for an initial six menopausal hormone products, including removal of breast-cancer risk statements from the boxed warning. The location of a warning in a label changed; the agency did not declare breast cancer impossible or approve HRT to prevent it. For example, the updated BIJUVA label still lists a history of breast cancer among contraindications and discusses breast risk in Warnings and Precautions.²³

Read the boxed-warning chronology for the regulatory sequence. This article answers a different question: how to interpret the clinical evidence and the current label together. A single screenshot of the old or new box is incomplete without the product’s indication, contraindications and remaining warning text.

Make the consultation population-specific

Bring whether the uterus is present, the exact symptom or labeled goal, personal and family cancer history, current screening context, and the proposed product/formulation. Ask the clinician which trial or guideline best matches the situation and where the remaining uncertainty lies. The family-history guide and post-cancer guide address those different histories directly; neither should be conflated with average-risk WHI estimates.

The hormone-therapy hub can help identify a practice to ask. A useful answer names the studied regimen, the relevant population, the outcome and the current exact-product label—without reducing a nuanced evidence record to “HRT causes cancer” or “FDA says there is no risk.”