Two companies presented fresh evidence at a Barcelona lung conference on September 8 that AI can sharpen how drug trials are measured.
Brainomix and Endeavor BioMedicines applied automated CT scoring to scans from ENV-IPF-101, a small phase 2a study of taladegib. The trial was randomized and placebo-controlled, and it lasted 12 weeks. Taladegib targets the hedgehog pathway, a signaling route tied to fibrosis. Earlier readouts showed the drug improved breathing measures.
The imaging tool, Brainomix 360 e-Lung, carries FDA clearance and a CE mark. It maps lung structure and disease burden from chest CTs without manual tracing.
In this analysis, taladegib came out ahead of placebo on lung volume, the amount of visible interstitial disease, and total scar tissue. The gaps were statistically significant.
The sponsors argue the readout matters because standard endpoints have limits. Forced vital capacity can miss biological changes that imaging catches, especially as newer background drugs flatten decline curves.
If regulators warm to such measures, developers could run leaner studies and get clearer signals from fewer patients. Brainomix said the outcome clears the way for more joint work with Endeavor on taladegib.
IPF slowly stiffens lung tissue, has few therapies, and carries a grim prognosis, so any tool that accelerates better trials is a clinical plus.
