PIANOMust knowUC and CD · Pregnancy
Pregnancy and Neonatal Outcomes After Fetal Exposure to Biologics and Thiopurines Among Women With Inflammatory Bowel Disease
Mahadevan U, et al.
Gastroenterology 2021 · PMID 33227283 (PubMed, opens in a new tab) · doi 10.1053/j.gastro.2020.11.038 (full paper, opens in a new tab)
| Outcome | Estimate (confidence interval) | P |
|---|---|---|
| Any infant infection in year 1 · adjusted odds ratio | ||
| Biologic (primary) | 0.92 (95% CI 0.70–1.20) | |
| Thiopurine (primary) | 0.90 (95% CI 0.64–1.28) | |
| Combination (primary) | 0.93 (95% CI 0.66–1.32) | |
| Preterm birth, The only independent risk factor for infection | 1.73 (95% CI 1.19–2.51) | |
| Spontaneous abortion before 20 weeks · hazard ratio | ||
| Biologic | 1.20 (95% CI 0.50–2.90) | |
| Thiopurine | 0.96 (95% CI 0.28–3.33) | |
| Combination | 0.96 (95% CI 0.28–3.33) | |
| Active disease | 3.41 (95% CI 1.51–7.69) | |
| Prior spontaneous abortion | 2.17 (95% CI 1.05–4.49) | |
Every drug interval crosses 1; the intervals that sit clear of 1 belong to active disease, prior spontaneous abortion and preterm birth.
Source: Mahadevan 2021, full text (PMC7956164), Results: Pregnancy outcomes and Infections. The thiopurine and combination hazard ratios for spontaneous abortion are printed identically in the paper.
Structured abstract
BackgroundLack of safety data had led some women with IBD to stop biologics or thiopurines in pregnancy, with health repercussions for mother and child.
MethodsProspective, observational, multicentre cohort study (30 US centres, 2007–2019); not randomised. 1,490 patients. Primary endpoint: Congenital malformations, spontaneous abortion, preterm birth, low birth weight and infant infections in the first year of life.
ResultsPrimary result. No increase in any primary outcome with biologic, thiopurine or combination exposure. Adjusted odds of any infant infection in year 1: biologic 0.92 (95% CI 0.70–1.20), thiopurine 0.90 (0.64–1.28), combination 0.93 (0.66–1.32).
Take-homeBiologics and thiopurines, alone or together, did not increase pregnancy or infant harms; active disease did, so continue effective therapy through pregnancy.
| Patients | 1,490 |
|---|---|
| Population | Completed pregnancies in women with IBD (1712 enrolled; singleton pregnancies only); 1431 live births, with 1-year infant outcomes in 1010. |
| Comparison | In utero exposure to thiopurines (242), biologics (642) or both (227) vs no exposure to either (379) |
| Drugs | infliximab, adalimumab, certolizumab pegol, golimumab, vedolizumab, ustekinumab, natalizumab, azathioprine, mercaptopurine |
| Classes | Anti-TNF, Anti-integrin, Anti-IL-12/23, Thiopurine |
| Published | Mahadevan U, et al. Gastroenterology 2021;160:1131–9. |
Key points
- Maternal disease activity, not medication, was linked with harm: active disease and spontaneous abortion HR 3.41 (95% CI 1.51–7.69); preterm birth and infant infection OR 1.73 (1.19–2.51).
- Women not on a biologic or immunomodulator were more likely to flare during pregnancy; UC was more often active and more often untreated than Crohn's disease.
- Infants had detectable drug at birth for all biologics except certolizumab pegol, mostly above maternal levels; birth drug level was not associated with infection.
- No differences in developmental milestones in the first year, or in a subset followed to 48 months.
- Later PIANO analyses: no increase in adverse outcomes with vedolizumab (66) or ustekinumab (47) exposure (PMID 37796648), and no differences in serious infection, malformations or developmental delay by medication in 438 children followed for at least 4 years (PMID 41789805).
Limitations
- Observational; outcomes largely self-reported by mothers; disease activity measured with clinical scores only.
- US cohort with high income and education, which may limit generalisability of developmental data.
- Few exposures to newer drugs in the main analysis; small molecules (JAK inhibitors, S1P modulators) were not studied.
Commentary
PIANO is the main reason IBD guidelines now advise continuing biologics and thiopurines throughout pregnancy rather than stopping anti-TNF therapy in the third trimester. The risks that matter come from active disease and preterm birth, not from drug exposure.
How to use it: plan pregnancy in remission and keep effective therapy going rather than stopping it early. Because biologics cross the placenta, tell the paediatric team about exposure; infant drug clearance and live vaccines are covered in the newborn drug-level study. Intestinal ultrasound in pregnancy can help confirm that disease is truly quiet.
Data for JAK inhibitors and S1P modulators in pregnancy remain limited; avoid them in pregnancy unless no alternative exists, following current guidance.
Journal club
Design
Prospective, observational, multicentre cohort study (30 US centres, 2007–2019); not randomised. 1,490 patients.
Primary result. No increase in any primary outcome with biologic, thiopurine or combination exposure. Adjusted odds of any infant infection in year 1: biologic 0.92 (95% CI 0.70–1.20), thiopurine 0.90 (0.64–1.28), combination 0.93 (0.66–1.32).
Validity and applicability
- Observational; outcomes largely self-reported by mothers; disease activity measured with clinical scores only.
- US cohort with high income and education, which may limit generalisability of developmental data.
- Few exposures to newer drugs in the main analysis; small molecules (JAK inhibitors, S1P modulators) were not studied.
Where it sits now
Biologics and thiopurines, alone or together, did not increase pregnancy or infant harms; active disease did, so continue effective therapy through pregnancy.
- Anti-TNF levels in newborns (2016): At birth, infliximab levels in the newborn are about twice the mother's and adalimumab levels are similar or slightly higher; both take months to clear (infliximab longest), which matters for live vaccines; combination with a thiopurine raised infant infection risk.
- Intestinal ultrasound in pregnancy (2025): In pregnancy, inflammation seen on intestinal ultrasound was associated with preterm birth, low birth weight and pre-eclampsia independently of symptoms and calprotectin; consider ultrasound to monitor IBD antenatally.
Questions for discussion
- This is an observational cohort. How could confounding by disease severity bias the drug comparisons, and in which direction?
- How would you counsel a woman in remission on an anti-TNF who wants to stop in the third trimester to reduce infant exposure?
- What does PIANO not tell you about JAK inhibitors or S1P modulators in pregnancy?