Antifibrotic dose cuts and stops were common in IPF, and progression common either way

Design

Retrospective analysis of a national registry and a health-system record

Population

1,572 patients with IPF starting nintedanib or pirfenidone in the US

Primary outcome

Composite progression (FVC fall of 10% or more, hospitalisation, transplant or death)

Effect

Registry: stop HR 1.51 (1.04–2.18), reduce HR 1.17 (0.90–1.52); EHR: no association

Two US data sources — a multicentre registry (545 patients) and a health-system record (1,027 patients) — followed people with IPF who started nintedanib or pirfenidone. Only 64% and 59% stayed on full dose; the rest reduced or stopped for three months or more, usually because of side effects.

Progression, defined as a 10% or greater fall in FVC, hospitalisation, transplant or death, was frequent in every group. In the registry, discontinuation was associated with faster progression (HR 1.51, 95% CI 1.04–2.18) while reduction was not clearly different (HR 1.17). In the health-system data, neither reduction nor stopping was associated with a difference.

The inconsistency between sources, and the possibility that sicker patients both tolerate the drugs worse and progress faster, limit conclusions. A reasonable reading is that a reduced dose that a patient can stay on is better than stopping, but that current antifibrotics leave much progression untouched.

• Expect dose reduction or interruption in about four in ten patients starting antifibrotics.
• Consider a reduced dose before stopping when side effects limit treatment.
• Manage diarrhoea and nausea actively to keep patients on therapy.
• Discuss early that progression often continues despite treatment, and plan transplant assessment where appropriate.

Why it matters

Tolerability, not efficacy, decides whether many IPF patients receive the drug at all.

Don't overread it

This was observational, and the two data sources disagreed; it cannot show that dose reduction is as effective as full dose.

The statistics, in plain English

A hazard ratio of 1.51 for stopping in the registry means about 50% faster progression, but the health-system data showed no such effect (HR 0.94, 0.66–1.36). When two sources disagree, neither result should be taken as settled.

Read the rest in the app

You have read your two free briefings this month. The app carries all 27 specialties, every morning, free — and this finding is waiting in it.

Get Daily Doctor — free

Scan to keep reading on your phone. No account needed to start.