Ophthalmology: Exp-AI driven biomarker strategy identifies responder subgroups in persistent ocular allergic inflammation.

This case study highlights how Ariana Pharmaceuticals’ KEM® Explainable AI technology was used to uncover relevant biomarker-defined patient subgroups from a Phase II ophthalmology clinical trial conducted by Fovea Pharmaceuticals (later acquired by Sanofi).
The study analyzed data from a randomized, double-blind, six-arm clinical trial evaluating FOV 1101 in combination with another therapeutic agent for patients with persistent allergic conjunctivitis. Although only two efficacy endpoints were statistically significant across the overall study population, KEM® identified responder subgroups with markedly improved treatment outcomes.
Using comprehensive clinical and biomarker data from 724 patients characterized by 165 patient descriptors across 12 clinical endpoints, KEM® enabled identification of patient populations most likely to benefit from treatment while simultaneously revealing the endpoints that best captured therapeutic efficacy.

- ● Identified five clinically actionable biomarker-defined patient subgroups associated with enhanced treatment response.
- ● Discovered 37 significant efficacy endpoints within these responder populations, compared with only two significant endpoints observed in the overall trial population.
- ● Supported biomarker-driven patient enrichment strategies for future ophthalmology clinical trials.
- ● Contributed to the scientific value of the FOV 1101 development program and showcased the role of KEM® in precision medicine and clinical decision-making.

- ●Identify biomarker-defined patient subgroups with improved response to FOV 1101 combination therapy.
- ● Discover clinically meaningful efficacy endpoints capable of capturing treatment benefit.

● Phase II randomized, double-blind, six-arm clinical study.
● Patients with persistent allergic conjunctivitis receiving FOV 1101 combination therapy.
● 724 patients included in the analysis.
● 165 demographic, clinical and biological descriptors evaluated for each patient.
● 12 predefined efficacy endpoints assessed across treatment groups.

- ● KEM® Explainable AI explored complex interactions between patient characteristics and clinical outcomes.
- ● Association-rule mining was used to identify biomarker combinations predictive of therapeutic response.
- ● Patient populations were stratified into biologically meaningful responder subgroups.
● Treatment efficacy was reassessed within each subgroup to identify endpoints demonstrating clinically relevant benefit.
● KEM®’s interpretable results were leveraged transparent interpretation of subgroup-specific treatment effects.

● The Ariana’s KEM® platform identified five clinically relevant patient selection biomarkers, defining distinct responder populations.
● While conventional statistical analysis identified only two significant efficacy endpoints across the entire study population, KEM® uncovered, across these responder subgroups, 37 additional efficacy endpoints revealing therapeutic benefits that were not apparent in the unstratified population.
● The Ariana Pharma’s analytical results played an important role during the strategic discussions preceding Sanofi’s acquisition of Fovea Pharmaceuticals.

● In October 2009, Sanofi announced the acquisition of Fovea Pharmaceuticals for up to €370 million (approximately US$539–541 million), including milestone payments.
● In its official announcement, Sanofi highlighted FOV 1101, the Phase II program for persistent allergic conjunctivitis, together with Fovea’s ophthalmology pipeline and proprietary discovery platform, as strategic assets supporting the acquisition.
● According to Pierre Belichard, Fovea’s former COO, the KEM® Clinical analyses were “instrumental in the outcome of discussions” during the M&A process that preceded the acquisition (quoted on Ariana Pharmaceuticals’ presentation). This statement reflects the perspective of a senior Fovea executive and complements the publicly announced transaction.