High-dose steroid pulses did not beat a lower dose in IPF flare-ups

When someone with idiopathic pulmonary fibrosis (IPF) suddenly gets much worse over days — an acute exacerbation — the short-term outlook is grim, and doctors reach for corticosteroids almost reflexively. Yet guidelines only weakly recommend steroids and do not say which dose or for how long [s1]. A new randomised trial set out to test the single most common habit: hitting the flare with very high "pulse" doses of intravenous methylprednisolone. The answer was that the big dose did not save more lives than a far smaller one [s1].

Why this question matters

IPF scars the lungs progressively, and an acute exacerbation is often the event that kills. Pulse methylprednisolone — gram-level daily doses — is used worldwide on the assumption that more steroid means more anti-inflammatory firepower against a sudden decline. But that assumption had little direct evidence behind it, and high-dose steroids carry real risks of infection, high blood sugar and muscle weakness in already fragile patients [s1]. The trial, called SAFER, was designed to find out whether the aggressive regimen actually earns its keep.

What the trial did

SAFER was a randomised, open-label, multicentre trial at eight South Korean university hospitals, registered as NCT04996303 [s1][s2]. Adults admitted with an acute exacerbation of IPF were assigned 1:1 to one of two regimens: intravenous pulse methylprednisolone at 10 mg/kg for 3 days, or a non-pulse regimen of 1 mg/kg for 7 days [s1]. Both groups then followed the same protocolised taper down to maintenance low-dose prednisolone, so the trial isolated the effect of the initial blast of steroid rather than the drug itself [s1]. The primary outcomes were all-cause mortality at 28 days and at 90 days [s1].

From July 2021 to April 2024, 100 patients were randomised — 51 to the pulse regimen and 49 to the non-pulse regimen [s1]. Secondary outcomes included intensive care admission, use of mechanical ventilation or extracorporeal membrane oxygenation (ECMO), the combined endpoint of death or lung transplantation, and adverse events [s1].

What it found

The high-dose pulse regimen did not lower deaths. At 28 days, 10.0% of the pulse group had died versus 14.6% of the non-pulse group — a risk difference of -4.6 percentage points, with a 95% confidence interval running from -18.8 to 9.1 (p=0.549) [s1]. At 90 days, mortality was 24.5% with pulse dosing and 29.8% without, a risk difference of -5.3 percentage points (95% CI -23.1 to 12.6; p=0.648) [s1].

Those point estimates numerically favour the pulse regimen, but the confidence intervals are the story: they are wide enough to be compatible with a meaningful benefit, no effect, or a meaningful harm [s1]. In other words, the trial could not distinguish the two strategies. When the investigators repeated the 90-day analysis treating lung transplantation as a competing event using a Fine-Gray model, they again found no difference [s1]. Secondary outcomes — ICU admission, ventilation or ECMO, and the death-or-transplant endpoint — were not demonstrably different between the groups either [s1].

How to read a null result

It is tempting to look at "24.5% versus 29.8%" and conclude the pulse dose helped. That reading is not supported. With only 100 patients, the trial was small, and a difference this size could easily arise from chance alone — which is exactly what a p-value of 0.648 signals [s1]. The honest summary is that SAFER did not show the aggressive regimen to be superior, not that it proved the two identical. Equally, the data give no license to assume the high dose is safe; a larger trial could still reveal either a modest benefit or the infection and metabolic harms that high-dose steroids are known to cause [s1].

What SAFER does do is undercut the reflex. If gram-level pulses are not clearly better than a gentler course for the outcome that matters most — survival — then the burden shifts to those who favour them to show the extra steroid is worth its risks [s1]. For a condition where clinicians have long acted on habit rather than evidence, a well-conducted trial that finds no advantage to the harder-hitting option is itself a useful result.

What to watch

The authors frame SAFER as a single multicentre trial in one country, and its modest size is its central limitation [s1]. Whether guideline writers revise their weak corticosteroid recommendations will likely wait for larger or pooled data. For now, the practical takeaway is narrow but real: there is no trial evidence that pulse-dose methylprednisolone rescues more patients from an IPF flare-up than a lower-dose regimen does.

Sources

• [s1] Pulse versus Nonpulse Corticosteroid Therapy in Acute Exacerbation of Idiopathic Pulmonary Fibrosis: A Multicenter, Randomized Clinical Trial (SAFER). American Journal of Respiratory and Critical Care Medicine, 27 September 2026. https://doi.org/10.1093/ajrccm/aamag529
• [s2] Efficacy of Steroid Pulse Therapy in Acute Exacerbation of Idiopathic Pulmonary Fibrosis (AE-IPF) Admitted in ER. ClinicalTrials.gov, NCT04996303. https://clinicaltrials.gov/study/NCT04996303