After reporting the first patient dosed in September 2025, early clinical data from a small open-label study indicate that Enveda’s oral candidate, discovered via the company’s PRISM AI platform, ENV-294, produced rapid reductions in disease severity and itch in patients with moderate-to-severe atopic dermatitis, alongside a safety profile without serious adverse events.
Colorado-based Enveda is a biotechnology company focused on systematically exploring natural chemistry for drug discovery, addressing a long-standing bottleneck where most molecules in nature remain uncharacterized due to the difficulty of isolating, identifying, and testing them.
Founded in 2019, the company has built a pipeline of more than a dozen candidate molecules, including programs in inflammatory bowel disease and metabolic conditions, with several already advancing toward clinical development. Enveda’s approach goes beyond pharmaceuticals to areas such as agriculture and materials. The company has raised a total of $517 million in funding, including last year’s $150 million Series D round.
Image credit: Enveda
A Multi-Pathway Approach to Atopic Dermatitis
ENV-294 is designed to treat inflammation by changing how immune signals are produced at the source, rather than blocking specific signals later like existing eczema drugs.
It is a non-degrading variant (referred to as a LOCKTAC), which means it does not remove proteins from the cell but instead stabilizes protein interactions to shift immune signaling. Biomarker data suggest activity across multiple immune pathways (Th1, Th2, Th17), indicating a broader modulation of inflammation compared to therapies that target a single cytokine axis.
Enveda plans to advance ENV-294 into Phase 2a trials in atopic dermatitis and asthma, with a Phase 2b study targeted for mid-2026.
During the Phase 1b trial, nine adult patients were treated with ENV-294 once daily at 800 mg for 28 days, followed by a 14-day observation period. The study included patients with prior systemic therapy exposure.
According to the release, clinical outcomes showed rapid and sustained reductions in disease severity, with responses emerging within the first week and continuing to improve even after treatment cessation. Most patients reached clinically meaningful response thresholds, and a substantial proportion achieved high levels of skin clearance, including cases of complete or near-complete resolution.
AI-Enabled System for Natural Product Discovery
Enveda’s platform combines a large natural-product database, AI models that interpret chemical data, and high-throughput lab testing. It turns plants and other natural materials into searchable, testable drug candidates in three steps:
- Organize — build a map of nature’s chemistry
Large numbers of plant samples are collected and a database is created linking plants, their molecules, and known disease associations. Using mass spectrometry, these samples are scanned to detect all chemical compounds within them, including many that have not been previously described. The result is a structured library where molecules from nature can be searched like entries in a database. - Translate — figure out what the molecules are
Instead of isolating and analyzing each molecule individually, a foundation model trained on large datasets of chemical signals, PRISM, is used to predict molecular structures directly from mass spectrometry data. This shifts analysis from a slow, manual process to one that can handle thousands of molecules at once. - Apply — test what the molecules do
An automated lab system screens these molecules in biological assays to evaluate behavior, for example, whether specific disease-relevant pathways are affected. This links chemical structures to potential therapeutic effects at scale.
To date, the platform has produced more than a dozen development candidates, three of which—ENV-294, ENV-308 (obesity), and ENV-6946 (inflammatory bowel disease)—have entered clinical trials.
Topic:
AI in Bio