Starting an Aromatase Inhibitor: Bone Protection Now or Only If Density Falls
A single patient about to start a drug that will help her cancer and hurt her bones at the same time. The disagreement is whether to protect the bone from the first dose or wait for her own numbers to show the damage first.
Yolanda P., a 57-year-old woman, was diagnosed with early-stage, estrogen-receptor-positive breast cancer four months ago and completed her lumpectomy and radiation without complication — about as good as this diagnosis gets, in her oncology team's phrasing. She is now facing five to ten years on anastrozole, which will suppress the small amount of estrogen her body still produces ten years past menopause and meaningfully lower her recurrence risk. She has never smoked and has no personal or family history of fracture.
The bone conversation is in this visit rather than a later one because aromatase inhibitors do not lower estrogen gradually. They remove the last real hormonal protection her skeleton has, and the effect starts from the first dose. Her baseline DXA, ordered before starting anastrozole per the standard workup, came back normal — and the specific values are what decide this. Her femoral neck sits at −0.6 and her lumbar spine at −0.9, both above the −1.0 line that guideline-based stratification uses to separate upfront treatment from monitoring, and she carries no other major risk factor to push her across it. ABCSG-18 demonstrated that denosumab given alongside an aromatase inhibitor reduces fracture incidence rather than merely preserving a density number, which is a real result and a meaningful one; ASCO and related guidance nonetheless base the upfront-versus-monitor decision on baseline T-score and clinical risk factors rather than on aromatase-inhibitor exposure alone. That places her outside the group in which upfront antiresorptive therapy has been shown to earn its risks — a statement about where the evidence currently reaches, not a prediction that her bone will hold.
Protecting the bone before the numbers show damage
I'd start denosumab now, alongside the anastrozole, rather than waiting for a follow-up DXA. Aromatase inhibitors remove the last real estrogen protection her bone had left, starting from the first dose — and ABCSG-18's trial data showed denosumab given alongside an aromatase inhibitor actually reduced fracture incidence, not just preserved a BMD number on a scan.
Her baseline DXA is normal, though — not osteopenic, nothing borderline. A reactive approach, treating only if a follow-up scan shows meaningful decline, avoids putting a genuinely low-baseline-risk patient on years of antiresorptive therapy she might not need, with its own real if uncommon risks.
I'm not arguing the fracture-reduction data is wrong — I'm arguing it doesn't automatically mean every AI patient regardless of starting bone density should get treated upfront, which isn't what most guidelines currently recommend either.
That's actually the right frame — this isn't "treat everyone on an AI" versus "treat no one," it's risk stratification. ASCO and related guidance base the upfront-versus-monitor decision on baseline T-score and clinical risk factors, not on aromatase-inhibitor use alone.
Her baseline T-scores are both above −1.0 with no other major risk factor — by that stratification, monitoring with a repeat DXA at one to two years, rather than starting denosumab today, is the guideline-supported path, with the plan explicitly revisited if that scan shows the decline the drug is expected to cause.
Agreed: hold denosumab for now given her normal baseline DXA and low clinical risk, start calcium and vitamin D immediately, and schedule a repeat DXA at one to two years rather than waiting for the full course of anastrozole to run before checking.
Not fully settled: whether one year or two is the right interval for that repeat scan — the oncologist preferred the shorter interval given how directly and quickly aromatase inhibitors are known to affect bone, the endocrinologist felt two years matched the guideline's own typical monitoring cadence; the one-year interval was chosen for this patient specifically, without either physician treating it as the fixed answer for every similar case.