Many powerful biologic drugs are antibodies, and a quirk of biology means they readily cross the placenta, raising safety questions for pregnant patients. Researchers report a design tweak that sharply reduces that transfer.
Scientists at the University of Oslo and Oslo University Hospital, led by Jan Terje Andersen, found that fusing IgG antibodies to albumin significantly reduces their transport across the placenta while keeping the drugs’ long-lasting effectiveness. The findings appeared in Science Immunology.
Why antibodies cross the placenta on purpose
This is not a leak. It is an evolved system, and understanding that is what makes the solution possible.
Newborns cannot make their own antibodies for months, so the placenta actively transports maternal IgG to the fetus, giving the infant a working immune repertoire at birth. It is why maternal vaccination protects newborns against pertussis and RSV.
The transport is performed by a receptor called FcRn, which binds IgG and ferries it across. The system cannot distinguish protective maternal antibodies from therapeutic ones — a monoclonal antibody drug looks like IgG because it is IgG, so it is transported with the same efficiency.
The clinical problem
Monoclonal antibodies treat many chronic conditions common in women of reproductive age — cancer, autoimmune diseases and migraine among them.
That creates a recurring dilemma. A woman with severe inflammatory bowel disease, rheumatoid arthritis or multiple sclerosis who becomes pregnant faces a choice between continuing a drug that reaches her fetus at concentrations that can exceed her own by late pregnancy, or stopping and risking a disease flare — which carries its own risk to the pregnancy.
The consequences are not hypothetical: infants exposed in utero to certain immunosuppressive antibodies have measurably altered immune function in early life, which affects decisions about live vaccines.
The albumin trick
The insight exploits a specific asymmetry. FcRn binds both IgG and albumin — it is the receptor keeping both proteins in circulation for weeks rather than hours — but it binds them at different sites, and the placenta transports IgG while largely leaving albumin behind.
Fusing the two creates a molecule that still engages FcRn, retaining the long half-life that makes antibody drugs practical, while apparently failing to satisfy whatever the placental transport machinery requires.
That is an unusually elegant piece of protein engineering: keeping the beneficial consequence of FcRn binding while losing the unwanted one, by changing which part of the receptor interaction dominates.
What was tested
The approach was evaluated in mouse studies, in ex vivo human placental tissue, and in a test model of fetal/neonatal alloimmune thrombocytopenia — substantially reducing fetal exposure and adverse effects.
The human placental tissue work matters most. Placental transport differs meaningfully between species, so a mouse result alone would be weak evidence. Demonstrating reduced transfer across actual human placental tissue addresses the obvious objection directly.
The chosen disease model is also apt. Fetal/neonatal alloimmune thrombocytopenia is caused by maternal antibodies crossing the placenta and destroying fetal platelets — a condition where placental transfer is the mechanism of harm, making it the cleanest possible test.
Why keeping the half-life is essential
An obvious alternative exists: engineer antibodies that do not bind FcRn at all, which would stop placental transport completely.
That fails for a different reason. FcRn binding is what gives antibodies their long half-life — without it, a therapeutic antibody would be cleared within days rather than weeks, requiring impractically frequent dosing.
The albumin fusion threads that needle, which is why it is more interesting than simply abolishing the interaction.
What it would mean
“We now have the opportunity to develop a new generation of biologic medicines that combine long-lasting efficacy with improved safety during pregnancy,” the researchers said.
Pregnant women are systematically excluded from drug trials, so evidence about biologics in pregnancy comes largely from registries and accumulated clinical experience. A drug designed from the outset to minimise fetal exposure would change that conversation.
The trial exclusion problem underneath this
The reason this engineering problem matters so much is that the clinical evidence it would replace largely does not exist.
Pregnant women have been systematically excluded from drug trials for decades, on reasoning that was protective in intent and produced the opposite result. Excluding them from research does not stop them needing medication; it means they receive drugs whose effects in pregnancy were never studied.
What fills the gap is registries, case series and accumulated clinical experience — useful, slow to accumulate and rarely definitive. Clinicians end up making decisions from indirect evidence, and patients end up choosing between an untreated disease and an inadequately characterised drug.
A biologic designed from the outset to minimise placental transfer changes the terms of that decision without requiring the trial evidence to catch up — which is why an engineering solution has appeal in a field where the evidentiary problem has proven so resistant to fixing directly.
The work is preclinical, and applying it means re-engineering existing molecules and testing them as new products rather than modifying approved drugs. That is a substantial development programme per drug — but it points to a practical route to making an entire class safer in pregnancy. Early research; not medical advice.