# Sanofi & Pangaea to use AI for finding rare lung disease patients

> Sanofi and Pangaea Data's AATD collaboration shows where pharma AI may create near-term value: finding undiagnosed patients already visible in clinical records. In rare disease, patient identification can create patient value and market expansion at the same time.

URL: https://www.ch-healthtech.com/insights/sanofi-pangaea-data-partnership-detect-underdiagnosed-genetic-disease
Markdown: https://www.ch-healthtech.com/insights/sanofi-pangaea-data-partnership-detect-underdiagnosed-genetic-disease.md
Published: 2026-06-25
Author: Christian Hein
Tags: activity/partnership-alliance, function/commercial-strategy, technology/digital-health, industry/large-pharma, technology/artificial-intelligence, therapeutic_area/rare-diseases-orphan-drugs

---


As many as nine in ten people in the US with a serious inherited condition that can destroy lung tissue are never tested for it. This week, Sanofi and Pangaea Data announced an AI collaboration to help find them.

The condition is Alpha-1 Antitrypsin Deficiency. It can drive early emphysema and COPD, and patients often wait five to eight years from first symptoms to diagnosis. By then, much of the lung damage may already be irreversible.

Here is what should change how pharma thinks about AI.

Many of these patients are already visible in the electronic health record. The breathing complaints. The repeat visits. The COPD label that was never questioned. The signal sits in the chart, unread, for years.

Sanofi's collaboration with Pangaea Data points a model at that unread signal. It uses structured data and clinical notes to surface people who may qualify for AATD evaluation, inside existing clinical workflows.

The drug discovery deals get the headlines. Molecule generation is a long bet, often measured in many years, and most candidates still fail.

Patient identification runs on a different logic. The condition exists. The patients exist. Standard therapies already exist, and Sanofi has an investigational AATD therapy moving through development. The missing step is often connecting the right patient to the right evaluation early enough.

There is a commercial truth here worth saying plainly. In rare and underdiagnosed diseases, better patient finding can create real patient value and real market expansion at the same time. That alignment is unusual, and pharma should (and increasingly does) pay attention to it.

My read: the quiet patient-ID deals may move pharma's P&L over the next three years more than many of the molecule-design partnerships getting ten times the attention. Particularly in rare diseases.

Find the patient first.