NEGATIVE RESULTS

aTyr's efzofitimod misses again on steroids in sarcoidosis lung subgroup

A 44-patient post hoc slice of the failed Phase 3 EFZO-FIT trial shows an FVC gain that is now anchoring aTyr's next registrational study, not the original steroid-sparing goal.

By AppliedXLJuly 15, 2026Trial NCT05415137

Executive Summary

aTyr is presenting a subgroup analysis from a Phase 3 trial that already missed its primary endpoint, using a lung-function signal in a small slice of patients to justify a redesigned follow-on study.
The company has shifted its registrational strategy from steroid reduction to lung function, and has already taken that new design to the FDA, ahead of any confirmatory data in the population it targets.
The improvement being highlighted comes from a post hoc, small-sample subgroup, while the same subgroup showed no separation from placebo on the metric the original trial was built to test.
No competing program pairs this target with this disease, and no other industry effort in the indication has reached the same late stage, leaving the mechanism's readthrough entirely dependent on aTyr's own data going forward.

The trial and the miss

EFZO-FIT (NCT05415137) enrolled 268 patients with pulmonary sarcoidosis and tested intravenous efzofitimod, an NRP2 agonist, against placebo, with the primary endpoint set as the change from baseline in mean daily oral corticosteroid dose at Week 48. The trial completed in 2025, and aTyr disclosed that the 5.0 mg/kg efzofitimod arm did not separate from placebo on that steroid-reduction endpoint across the 264-patient treated population, even as several secondary measures reached statistical significance, including KSQ-Lung score (p=0.0479) and complete steroid withdrawal with KSQ-Lung improvement (p=0.0196). NCT05415137+1

Probability of Success

Endpoint Met0%

Completes81%

Clinical Significance73%

Regulatory86%

The subgroup: a 44-patient slice

The poster now being presented narrows further, to a prespecified subgroup of 44 patients with restrictive lung disease, defined as FVC percent predicted of 80% or less with a normal FEV1/FVC ratio. In that subgroup, the placebo-adjusted Week 48 change in forced vital capacity for the 5.0 mg/kg arm was 123.8 mL by random coefficient regression model, alongside improvement trends on the King's Sarcoidosis Questionnaire-Lung score, general health score, fatigue scale, and a cough questionnaire. The magnitude of steroid reduction in that same subgroup, however, was similar to placebo, meaning the trial's actual primary endpoint still did not separate in the population aTyr is now emphasizing. aTyr

The regulatory pivot

Following what the company describes as recent interactions with the FDA, aTyr submitted a protocol in June 2026 for a new Phase 3 study restricted to patients with restrictive lung disease, using FVC as the primary endpoint and the KSQ-Lung score as the key secondary endpoint. Chief Executive Sanjay Shukla called the approach "a well-informed plan to demonstrate the potential for efzofitimod to improve the lives of these patients". The shift moves the registrational bar from a steroid-sparing claim, which EFZO-FIT could not support, to a lung-function claim drawn from a subgroup that was not the trial's powered analysis population. aTyr

The competitive and regulatory backdrop

Efzofitimod holds FDA Fast Track designation and orphan drug designation in the United States, European Union, and Japan for sarcoidosis, reflecting the disease's unmet need rather than any judgment on the drug's chances. No other industry trial pairs the NRP2 target with pulmonary sarcoidosis, and efzofitimod remains the only Phase 3 asset in that combination; other sarcoidosis programs in testing, including Novartis's canakinumab and MAS-825, Kinevant Sciences' namilumab, and Mallinckrodt's corticotropin, work through different mechanisms and none has produced a resolved same-target precedent for aTyr's platform. That leaves the new FVC-anchored trial as the mechanism's own test of whether the restrictive-subgroup signal generalizes, with no peer readout to lean on.

This analysis was produced using AI-assisted reporting systems, AppliedXL data, and official public records. These systems undergo editorial review, quality checks, and regular audits by human experts. Errors may still occur, as with any automated system. Always consult the linked primary sources. Read our AI Editorial Policy.