Chronic Suppressive Therapy in Bronchiectasis: PROMIS-I Against Its Own Twin
A frequent-exacerbator with chronic Pseudomonas colonization, whose two candidate suppressive therapies are backed by directly conflicting trial results from the same research program. The disagreement is whether PROMIS-I’s positive result still stands once its prematurely-terminated twin’s null result is taken seriously — and what to do about NTM-resistance stewardship before starting either option.
James K., a 58-year-old man, retired early last year after a heart attack scare that turned out, after a week of tests, to be nothing cardiac at all — but the scare was enough to finally push him to finish the woodworking shop in his garage he’d been planning for a decade, mostly for an excuse to spend Saturdays building furniture instead of watching golf. He was diagnosed with bronchiectasis six years ago and has never been tested for cystic fibrosis specifically, though a sweat chloride test done years earlier for an unrelated workup came back normal, which the team has treated as reasonable confirmation that his disease is non-CF in origin. Over the past year he has had three exacerbations requiring antibiotics — two oral courses and one that put him in the hospital for four days on IV therapy — more than he’d had in any prior year, and sputum cultures across all three have consistently grown Pseudomonas aeruginosa. He has never been treated for or screened for nontuberculous mycobacteria, a gap nobody had flagged as relevant until today’s conversation about starting a long-term suppressive drug raised it.
Both trials required either two exacerbations needing oral antibiotics or one needing intravenous therapy in the prior year; James had two oral courses and one intravenous, so he clears the entry criterion on both of its limbs rather than scraping past on one, which puts him squarely inside the enrolled population rather than at its edge. PROMIS-I found an annualized exacerbation rate of 0.58 on inhaled colistimethate against 0.95 on placebo — a rate ratio of 0.65, a 35% reduction, with improved quality-of-life scores alongside it. PROMIS-II was not a later replication but an identically-designed trial running in parallel; it hit the COVID-19 pandemic, was terminated early at 287 patients, and found rates of 0.89 against 0.89, a rate ratio of exactly 1.00. Its investigators attributed much of that to pandemic disruption of visit schedules and of how exacerbations were even ascertained — disruption that could plausibly mask a real effect rather than prove its absence. Whether James’s own case should be read against the trial that found benefit or the one that didn’t is not a question either trial alone answers for him.
Pulmonary clinic, discussing suppressive therapy
Start inhaled colistimethate. In PROMIS-I, patients with bronchiectasis and chronic Pseudomonas — James’s exact phenotype — dropped from an annualized exacerbation rate of 0.95 to 0.58, a rate ratio of 0.65, with better quality-of-life scores over twelve months. Three exacerbations in a year, all Pseudomonas-positive, is exactly the population that trial was built around.
PROMIS-I isn’t the whole story you’re citing it as. Its twin, PROMIS-II — same design, run at the same time, not a small pilot that failed to confirm a real effect — found rates of 0.89 against 0.89, a rate ratio of 1.00. I’ll concede the obvious counter before you make it: it was stopped early at 287 patients, so it is the smaller of the two and underpowered for it.
I’d start chronic low-dose azithromycin instead, which has its own real trial support in non-CF bronchiectasis. But before starting any chronic macrolide, we need a sputum NTM culture first. Undetected MAC or M. abscessus colonization exposed to years of low-dose macrolide monotherapy is exactly how a patient acquires macrolide resistance in an organism where losing that drug matters far more.
You’re right that PROMIS-II is a genuine non-replication rather than background noise — and fair to have conceded it was the truncated one.
But its own investigators pointed to COVID-era disruption of visit schedules and exacerbation ascertainment as a plausible confound — that’s real uncertainty, not proof of no benefit. Send the NTM screen either way, since it’s needed regardless of which drug we choose. Start colistimethate with an explicit six-month exacerbation-count check as a predefined stopping rule, and keep azithromycin as the fallback if it doesn’t work — or immediately, if the NTM screen comes back positive and rules out chronic macrolide therapy on its own.
Agreed: send the NTM sputum culture today, and start inhaled colistimethate with a six-month exacerbation-count check built in from the start.
Not agreed: what happens if colistimethate fails its six-month check and the NTM screen is negative. The pulmonologist sees azithromycin as the clear, automatic next step at that point; the stewardship physician is less certain, given azithromycin’s own long-term resistance-selection costs even outside the NTM-specific concern already addressed.