A 5-Year Risk of 4.1 Percent: Deciding Whether a Preventive Drug Is Worth Taking
She has never had cancer and may never get it. A calculated risk score clears the threshold for chemoprevention, but three different drugs clear it in three different ways, each trading a specific, real side-effect profile for a modest reduction in a risk that is already, honestly, a minority probability.
Eleanor V., a 47-year-old county clerk who has spent nineteen years processing other people's marriage licenses and property deeds with the same unhurried thoroughness, brought that same thoroughness to her own risk-reduction consultation — she had already calculated her own Gail-model score from a tool she found online before her appointment, and arrived wanting to understand not just the number, 4.1% over five years, but what taking a preventive drug for a disease she may never develop would actually cost her in exchange.
A 5-year risk of 4.1% clears the threshold most guidelines use to consider pharmacologic chemoprevention, but clearing a threshold for consideration is a genuinely different thing from a clear mandate to treat, and the three drugs available to her each trade a different toxicity for a similar, modest reduction in that already-modest baseline risk. Tamoxifen, the original chemoprevention agent, carries a real increased risk of endometrial cancer and venous thromboembolism, concerns that matter specifically because Eleanor still has an intact uterus and is only perimenopausal, putting her years inside the window where those risks accrue. Raloxifene, tested head-to-head against tamoxifen in the STAR trial (NSABP P-2) in postmenopausal high-risk women, showed slightly less invasive breast cancer risk reduction but a meaningfully better endometrial and thromboembolic safety profile, though it is only approved and studied for postmenopausal use — a real timing mismatch with Eleanor's current perimenopausal status. Anastrozole, tested for prevention in IBIS-II — MAP.3 is the parallel chemoprevention trial, and it tested exemestane, not anastrozole — showed a substantial relative risk reduction with a different toxicity profile entirely — joint symptoms and accelerated bone loss rather than uterine or clotting risk — but like raloxifene, its trial population was postmenopausal, and using it before menopause runs into both an efficacy question, since ongoing ovarian estrogen production would blunt an aromatase inhibitor's effect the way it does in a premenopausal patient, and a genuine label mismatch.
Breast risk-reduction clinic
I want to argue for starting her on tamoxifen, and I want to argue it rather than open by hedging. A 4.1% five-year risk is two and a half times the 1.67% consideration threshold, and she has a mother diagnosed at 58 and an aunt at 61 — this is not a borderline score. Tamoxifen is the one agent with prevention data in premenopausal and perimenopausal women, which is precisely the objection about to be raised against the other two. Waiting for menopause to make the drug menu tidier means declining to treat a risk that is accruing now, in the years she is actually at risk.
Then let me be equally direct from the uterine side, because you have picked the one agent my objection actually lands on. An intact uterus in a perimenopausal woman is the population where tamoxifen's endometrial cancer signal is concentrated, not a background consideration, and she has years of exposure ahead of her rather than a short tail. I'm not saying no. I'm saying that if tamoxifen is chosen, baseline and annual endometrial surveillance is built in from the start, and she is told at the outset that we are trading a modest reduction in one cancer risk for a real increase in another.
Which she may well accept — women take that trade every day, and framing it as disqualifying rather than as a trade is its own kind of paternalism. What I won't do is let it be presented as the safe conventional option and the other two as the compromises.
I'd push back on the premise both of you are sharing, which is that something has to be started. Her cycles have been irregular for eight months — she is not years from menopause, she is plausibly months from it, and that changes what deferral costs. Wait, and all three agents come into range: raloxifene and anastrozole stop being off-label, tamoxifen's endometrial risk falls with the uterine lining it acts on, and the STAR and IBIS-II data start describing her instead of approximating her.
So the most defensible near-term option is watchful waiting with a scheduled reassessment — not because her risk is low, but because a short, bounded wait converts three imperfect fits into three real choices.
That's not inaction — it's a deliberate, time-bound plan with a clear re-evaluation point, which is a genuinely different thing from declining to treat and moving on. The medical oncologist is right that risk accrues in the interval, and I don't have an answer to that beyond its being a short interval.
Agreed: no chemoprevention agent started today, with a structured reassessment at roughly twelve months or sooner if her cycles stop entirely, and all three drugs on the table depending on how her risk score and menopausal status look then. Not agreed: the medical oncologist accepted the plan without being persuaded by it, and asked that the record show she considers a bounded wait to be a real cost in a woman whose five-year risk is 4.1% rather than a neutral pause — a position she expects to revisit if Eleanor's cycles turn out not to stop on the timetable the deferral assumes.