HRT and future ovarian-cancer risk: the WHI regimen distinction
Long-term WHI follow-up found higher ovarian-cancer incidence with the studied oral estrogen-only regimen after hysterectomy; the separate oral estrogen-plus-progestin trial did not show a statistically significant increase. Neither estimates every modern HRT regimen.
The best-known WHI ovarian-cancer result is regimen-specific. In long-term follow-up of randomized trials, oral conjugated equine estrogen alone after hysterectomy was associated with higher ovarian-cancer incidence than placebo: 35 versus 17 cases, hazard ratio 2.04. In a separate trial of oral conjugated estrogen plus medroxyprogesterone among women with a uterus, the estimate was 75 versus 63 cases, hazard ratio 1.14, with a confidence interval including no difference. These data do not yield one risk number for all HRT patches or formulations.1
This page addresses future cancer incidence in women who had not been treated for ovarian cancer at enrollment. The question of hormone therapy after an ovarian-cancer diagnosis has different evidence and a different clinical team. It would be misleading to answer one question with the other’s study.
Why the two WHI arms should not be pooled into a slogan
The estrogen-alone trial enrolled 10,739 women with prior hysterectomy. The combination trial enrolled 16,608 women with a uterus. The first tested oral conjugated equine estrogens (CEE) versus placebo; the second tested oral CEE plus medroxyprogesterone acetate (MPA) versus placebo. Different eligibility and hormone combinations mean they are not a randomized patch-versus-pill or estrogen-alone-versus-combination comparison.1
At approximately 20 years of follow-up, the estrogen-alone ovarian-incidence hazard ratio was 2.04 with a 95% confidence interval of 1.14–3.65. The combination result was 1.14 with a 95% interval of 0.82–1.59. The latter interval includes 1.0, so the data do not establish an increase for that studied regimen. They also do not prove zero possible effect; uncommon cancer outcomes leave uncertainty even in a large trial.1
| Study arm | Population and oral regimen | Interpretation |
|---|---|---|
| Estrogen alone | Prior hysterectomy; CEE versus placebo | Increased ovarian incidence in this long-term follow-up |
| Estrogen plus progestin | Intact uterus; CEE plus MPA versus placebo | No statistically significant ovarian-incidence increase shown |
| Modern estradiol patch | Not directly tested by these two WHI assignments | Do not copy either oral-regimen estimate |
Put the result beside the decision actually being made
A consultation should identify the exact proposed hormone, route, progestogen if any, and reason for therapy, plus the person’s uterus and relevant cancer history. A family history or inherited-risk question belongs in an individual risk assessment; these WHI group results cannot classify it. The NCI hormone-therapy fact sheet is a useful primary public summary of regimen-dependent cancer questions.2
The WHI breast-cancer guide addresses a separate organ and endpoint; its numbers should not be swapped into this ovarian discussion. Ask a hormone-therapy provider to state which evidence actually matches the product under consideration, and which uncertainty remains when the product differs from the WHI oral regimens.
Rare outcomes need both counts and uncertainty
The hazard ratio of 2.04 for estrogen alone can sound more certain than the underlying ovarian case counts. Thirty-five versus seventeen cases over long follow-up are enough for a statistically significant signal in this analysis, but the confidence interval is wide. The ratio is about the rate of events over follow-up in groups under study conditions; it does not say that 2.04% of users developed cancer or that every individual’s risk doubled in a clinically transferable way. The counts and interval belong beside the estimate whenever it is quoted.1
The combination trial’s confidence interval includes 1.0. That means a clear increase was not demonstrated for the studied CEE-plus-MPA regimen; it does not make all hormone combinations protective or establish no possible risk. Distinguishing a finding from an absence of evidence is especially important for a less common outcome. A clinic marketing “safe because the combined trial was negative” would be claiming more than this analysis supports.1
Separate baseline history from treatment evidence
An inherited predisposition, family history, previous ovarian procedure, or other cancer history may affect a person’s baseline conversation. WHI’s randomization helps assess the studied treatments at a group level, but it is not a personalized risk model for those factors. The clinician can review the full history, the reason systemic therapy is being considered, and any alternatives that address the actual menopause symptom. NCI’s summary is useful because it keeps different hormone regimens and cancer outcomes separate.2
When reading a headline, check four fields: which population, which exact hormones, which cancer endpoint, and which follow-up interval. If one is missing, the claim may be too broad to guide a decision. This small audit also prevents an unrelated HRT-after-endometrial-cancer discussion from being mistaken for ovarian-incidence evidence.
Sources
- PubMed. Ovarian cancer incidence and mortality in long-term WHI hormone-trial follow-up. Randomized-trial follow-up reports separate oral CEE-alone and oral CEE-plus-MPA ovarian-cancer results over approximately 20 years. Accessed .
- National Cancer Institute. Menopausal hormone therapy and cancer risk. NCI contextualizes cancer-risk evidence by hormone regimen and population. Accessed .