The Birds Stayed: Treating Fibrotic Lung Disease That Outlasted an Antigen Removed Too Late
A woman with fibrotic hypersensitivity pneumonitis from years of pet bird exposure has already given the birds away. The disagreement now is whether antigen avoidance, the foundation of every HP guideline, still has anything left to offer once the disease has moved past the stage avoidance was ever going to fix.
C.W., a 66-year-old retired postal worker, kept finches and a cockatiel for close to twenty years, a hobby she describes as the one thing that made her small apartment feel less empty after her husband died. She rehomed all three birds eight months ago, on her pulmonologist's direct recommendation, once bird fancier's lung was first suspected — a harder decision than it might sound, and one she has mentioned more than once she still second-guesses on quiet evenings.
The avoidance came too late to stop what was already in motion. Serial HRCT shows honeycombing and traction bronchiectasis in a peripheral, basal-predominant pattern, and her FVC has fallen from 78% predicted a year ago to 64% predicted now — a relative decline of roughly 18%, comfortably past the 10% threshold INBUILD used to define progression despite appropriate management, and a decline that continued for the first several months after the birds left rather than plateauing the way early-stage inflammatory HP typically does once the antigen is removed. Whether she belongs in that trial's population or in the immunosuppression literature is not a framing question; on her own two numbers she is inside INBUILD's entry definition and was never inside a cohort selected for ongoing-decline-despite-avoidance at all. Precipitating antibodies to avian proteins remain positive, though that finding lags real exposure by months, and with hypertension on a single agent and a lifelong nonsmoking history there is little else left to blame for a decline this steep.
What the positive precipitins cannot tell the team is the thing they would most like to know: whether the antigen is still driving anything, or whether the disease has crossed into a phase that no longer needs her to be near a bird to keep progressing. She still keeps the empty birdcage in a closet rather than getting rid of it. Her daughter has started coming to appointments since the FVC drop was first flagged, mostly, C.W. says, to make sure she actually asks the questions she says she will and then doesn't.
Multidisciplinary ILD conference
Confirming and maintaining antigen avoidance is still the foundation here, and I'd add immunosuppression on top of it, not instead of it. Morisset and colleagues' retrospective comparison of mycophenolate and azathioprine in chronic HP found a significant improvement in DLCO — about 4% at one year — which I read as real evidence that an ongoing inflammatory component can still respond even once fibrosis has set in. I'll state the limit of that myself, because it matters here: FVC did not significantly improve in that cohort, and FVC is the number falling in front of us.
I'd weight her actual trajectory more heavily than the antibody titer. Her FVC kept falling for months after the birds were confirmed gone — that's not a pattern of ongoing antigen-driven inflammation responding to avoidance, that's the specific behavior INBUILD was built to identify: progressive pulmonary fibrosis independent of the original trigger. Its HP subgroup showed nintedanib slowed the rate of FVC decline regardless of whether avoidance had already happened.
I'd go further on the concession you just made, because I think it decides this. A gas-transfer improvement with no accompanying FVC signal, in a retrospective cohort with no placebo arm, is not evidence for the endpoint we are actually watching fall. Her honeycombing and sustained decline after confirmed avoidance put her closer to INBUILD's progressive-fibrosing entry criteria than to the patients that immunosuppression comparison was drawn from.
Her own numbers actually answer the question the two of you are debating in the abstract. INBUILD's entry criteria required a physiologic decline meeting a defined threshold despite appropriate management, in a fibrosing ILD pattern — and her FVC drop of 14 percentage points over a year, continuing well past confirmed antigen removal, sits directly inside that population, not at its margin. That's a stronger, more specific match to her than the immunosuppression cohort, which wasn't selected on ongoing-decline-despite-avoidance the way INBUILD was.
That doesn't rule immunosuppression out permanently — if nintedanib slows but doesn't halt her decline, mycophenolate remains a reasonable add specifically for whatever inflammatory component her positive precipitins might still represent. But starting with the drug whose trial population she most precisely matches is the more defensible first move.
Agreed: nintedanib started, with FVC and DLCO rechecked at three months to establish whether the rate of decline has genuinely slowed rather than assuming it from the drug's trial data alone.
Not agreed: whether mycophenolate should be added now, alongside nintedanib, rather than waiting to see nintedanib's effect first. The allergist would start both together given how much lung function she has already lost; the pulmonologist wants a clean read on nintedanib's own effect before adding a second drug that would make any further decline harder to attribute. Deferred to the three-month recheck.