For the families who have lived through a baby arriving weeks too soon, a finding from Stanford Medicine offers something rare: a new idea. Researchers there report that two familiar drugs, aspirin and lansoprazole, may delay labor far more effectively together than aspirin does alone. The study appeared online Oct. 21 in Science Advances.
Why the timing matters
Prematurity, meaning birth three or more weeks before the due date, is now the leading cause of death in children under 5 worldwide. Senior author Brice Gaudilliere, a professor of anesthesiology, perioperative and pain medicine, put the scale plainly: preterm birth “affects 10% of pregnancies worldwide.” It has been decades, he noted, since a new drug was found to prevent it.
The work builds on two decades of Stanford research into how the mother’s immune system shifts during pregnancy. Early on, her body learns to tolerate a fetus that is immunologically different from her. Late in pregnancy, a series of immune changes, including a swing toward pro-inflammatory signals, helps prepare her for labor. When this immune clock runs too fast and those signals switch on too soon, labor can start early.
Building a digital twin
The team built a cell-by-cell atlas of the immune system and used it to create “digital twins” of individual pregnant women, so drugs could be tested virtually. The system, called SIMCO, uses AI and machine learning. To train it, the researchers took blood samples from pregnant women and exposed the immune cells to many different drugs. “We measured responses from a quarter of a billion cells,” Gaudilliere said.
They then fed in longitudinal data from 63 pregnancies, linked to clinical outcomes, so the model could learn which cell patterns match which states and how patients differ from one another. “You’re making the prediction for one individual based on knowledge of millions of single cells from dozens of patients,” said co-lead author Jakob Einhaus, a postdoctoral scholar. After checking that SIMCO forecasts time to labor accurately, the team screened nine drug regimens chosen because they were already known to be safe in pregnancy. A drug-repurposing pipeline from Marina Sirota’s group at the University of California, San Francisco supplied them.
What the screen found
The aspirin and lansoprazole pair stood out. Low-dose aspirin alone has a modest ability to prevent premature birth, and lansoprazole is sometimes prescribed in pregnancy for acid reflux. “In the digital twin system, these two drugs in combination had a large effect on the delay of labor, 15 times higher than aspirin alone,” Gaudilliere said. The team was surprised by how strongly the model predicted it. In the combination, immune processes that help a pregnancy continue were enhanced, while those that trigger labor were inhibited. “So it all makes sense biologically,” he said.
Folic acid also slowed immunological progression toward labor. It is already recommended in pregnancy to lower the risk of spinal defects, and Einhaus said it was a surprise to see it have such profound immune effects.
Because the twins capture differences between patients, the method may also flag likely responders and non-responders. Enrolling only likely responders could let future trials answer questions with fewer participants.
On to patients
The team tested the pair in an animal model of an inflammatory state in pregnancy and found it delayed the onset of labor there. The U.S. Food and Drug Administration then approved a human clinical trial of the combination. Gaudilliere called it a rapid path from discovery work to testing in patients.
Researchers from UC San Diego and the AP-HP Sorbonne University Hospital Group in Paris also contributed. Funding came from sources including the Stanford Maternal and Child Health Research Institute, the March of Dimes Prematurity Research Centers at Stanford and UCSF, and the National Institutes of Health.
Sources
- ‘Digital twins’ of immune system in pregnancy identify drugs to prevent early labor, Stanford Medicine-led study finds
- Digital twin
- Developing an AI Factory Digital Twin with an Agent-Assisted Workflow | NVIDIA Technical Blog
Photo: LPS.1 (CC0)





