Premature Labor Treatment Under Review: The APOSTEL 8 Debate, Explained for Patients

Table of Contents

• Key Points
• What Is a "Correspondence" and Why Should Patients Read It?
• Background: Why Preterm Birth Treatment Matters
• Study Methods: How the APOSTEL 8 Trial Was Run
• Key Finding: More Steroid Courses Completed, Yet No Better Baby Outcomes
• The Big Questions About Steroid Timing and Dosing
• Single Babies Versus Twins: An Unanswered Question
• Overtreatment and Long-Term Worries About Steroids
• Future Directions: Rebuilding Preterm Birth Protocols From Evidence
• Limitations: What This Correspondence Cannot Prove
• Recommendations: Key Takeaways for Patients
• Other Corrections in the Same Notice
• Frequently Asked Questions
• Source Information

Key Points

• Atosiban delayed birth past 48 hours more often than placebo in APOSTEL 8, but newborn outcomes were not improved.
• Current risk assessment for threatened preterm birth needs urgent improvement because clinical practice results in overtreatment.
• The ideal time between steroid injection and birth is uncertain; evidence suggests benefits may extend up to 14 days.
• Whether atosiban works equally in twin pregnancies is unknown because subgroup differences were not statistically significant.
• Children from the APOSTEL 8 trial are being followed for 4 years to clarify long-term effects of steroid exposure.

What Is a "Correspondence" and Why Should Patients Read It?

A correspondence is a letter published in a medical journal. It is part of the normal give-and-take between scientists after a major study appears.

Readers may challenge how a study was interpreted. The original authors then respond in writing. These exchanges sharpen the scientific record. They also reveal where evidence is strong and where it is thin.

This particular correspondence answers letters by Loussert and colleagues and Zegarra and colleagues. The replies come from Larissa I van der Windt and Martijn A Oudijk, the research team behind the APOSTEL 8 trial. The note also contains an official "Department of Error" section that corrects two unrelated Lancet papers.

Background: Why Preterm Birth Treatment Matters

Preterm birth means a baby is born before 37 completed weeks of pregnancy. Babies born early can face breathing problems, feeding difficulties, and longer hospital stays. The earlier the birth, the higher the risk.

"Threatened preterm birth" is the warning sign that labor may be starting early. It is called "threatened" because not every patient who has contractions or cervical changes will actually deliver soon. Some go on to carry much longer.

When preterm birth seems possible, doctors use two main tools:

• Tocolytics (medications that slow or stop contractions) to buy time.
• Antenatal corticosteroids (steroid injections given to the mother before birth) that speed up the baby's lung development and reduce the risk of breathing problems and death after premature birth.

A standard corticosteroid course often consists of two doses given 24 hours apart, completed within 48 hours. The hope is that tocolytic treatment delays birth long enough for the full steroid course to work.

Study Methods: How the APOSTEL 8 Trial Was Run

APOSTEL 8 stands for "Atosiban versus placebo for threatened preterm birth." It was a multicenter randomized controlled trial—meaning many hospitals took part, and patients were randomly assigned to receive one of two treatments. Random assignment helps ensure the groups are comparable.

One group received atosiban. Atosiban is a tocolytic drug that blocks oxytocin receptors, the chemical "locks" that contractions must open. Blocking them relaxes the uterus.

The other group received a placebo (an inactive substance). Neither the patients nor their doctors knew which treatment was given during the trial.

The trial was designed as a pragmatic study, which means it tried to reflect routine real-world practice. That is why the researchers included a broad population rather than a narrowly selected one. Both singleton pregnancies (one baby) and twin pregnancies were allowed.

The study was published in The Lancet in 2025 (volume 405, pages 1004–1013). The overall conclusion from the trial's primary analysis applies to the entire study population, not just one subgroup.

Key Finding: More Steroid Courses Completed, Yet No Better Baby Outcomes

The researchers were careful to state that APOSTEL 8 was not designed to test how well corticosteroids work. Even so, one finding stood out.

Many more participants in the atosiban group completed a full corticosteroid course than in the placebo group. This makes sense: atosiban delayed labor longer, giving the two steroid doses time to be finished.

Here is the surprise. Completing the full steroid course did not lead to improved neonatal outcomes (better health results for newborns) compared with the placebo group.

The authors also reported that atosiban treatment resulted much more often in pregnancy prolongation beyond 48 hours compared with placebo. In other words, the drug genuinely delayed birth past the critical two-day window.

Yet that extra delay may not have been enough. The authors write that this prolongation "might be insufficient to achieve the full benefits of corticosteroids." Delaying labor is a means to an end. If the delay does not line up with the window when steroids help most, the baby may not benefit.

Something else matters for interpretation. Although a subgroup analysis suggested a possible difference in treatment effect between singletons and twins, this difference was not statistically significant. In plain language, the difference could easily have happened by chance. That is why the authors stress that the primary conclusion applies to the overall group.

The Big Questions About Steroid Timing and Dosing

The correspondents—and the APOSTEL 8 team themselves—raised several unresolved issues about how steroids are used. These issues may help explain why the atosiban group's extra steroid completions did not translate into healthier babies.

• Full course versus half course: Evidence supporting the benefit of a full course over a half course "remains sparse." Doctors need better proof that both doses are always better than one.
• Late gestational ages: The effectiveness of corticosteroids beyond 30+0 weeks of gestation (30 weeks and 0 days) is also poorly supported. Yet many patients threatening preterm birth are at or past this point.
• The administration-to-birth interval: The ideal amount of time between giving steroids and the baby being born remains uncertain. Evidence cited by the authors suggests the benefit window "might extend up to 14 days." If the interval matters that much, delaying birth by just over two days may not be enough.

These gaps matter for everyday care. A patient who receives atosiban, finishes her steroid course, and gives birth at 34 weeks is a medical success story on paper. Whether those steroids actually improved that baby's outcome is a separate question—and the answer is not yet clear from the evidence.

The researchers referenced a 2025 study by Melamed and colleagues in JAMA Network Open (volume 8, article e2511315) that examined the timing of antenatal corticosteroid administration and newborn outcomes. That study is part of the growing effort to pin down the optimal treatment window.

Single Babies Versus Twins: An Unanswered Question

Loussert and colleagues highlighted that the APOSTEL 8 trial included both singleton and twin pregnancies. Data on the use of tocolytics specifically in twin pregnancies are low, as the correspondents correctly note.

The APOSTEL 8 team agreed with that concern. They pointed to a 2022 Cochrane network meta-analysis by Wilson and colleagues (Cochrane Database of Systematic Reviews, article CD014978) on tocolytics for delaying preterm birth. In that analysis, differential outcomes between groups—meaning singletons compared with twins—have not been reported.

Why did APOSTEL 8 include twins then? Because it was a pragmatic trial meant to mirror real practice. Twins are common in preterm-birth clinics, and excluding them would have made the results less realistic.

The subgroup analysis hinted that twins might respond differently to treatment than singletons. But because the difference was not statistically significant, no firm conclusion can be drawn. For now, the honest answer is that we do not know whether atosiban works equally well—or equally poorly—in twin pregnancies.

Overtreatment and Long-Term Worries About Steroids

The correspondence makes a pointed criticism of current care: risk assessment for threatened preterm birth "needs urgent improvement," because current clinical practice results in overtreatment.

What does overtreatment mean here? Many patients who are labeled as having threatened preterm birth never actually deliver early. Yet they still receive tocolytics and steroids. That means some babies are exposed to potent medications without clear benefit.

Zegarra and colleagues raised concerns about the potential long-term adverse effects of antenatal corticosteroids in children who are born at full term—children who were exposed to steroids before birth but ended up being born on time anyway.

The APOSTEL 8 team cited a 2022 systematic review and meta-analysis by Ninan and colleagues in JAMA Pediatrics (volume 176, article e220483) that evaluated long-term outcomes associated with preterm exposure to antenatal corticosteroids. The full picture of how these steroids shape later child development is still emerging.

This is precisely why the trial has a planned 4-year follow-up. The researchers will track the children from APOSTEL 8 over four years to gain more insight into these long-term outcomes. That follow-up builds on a 2024 systematic review by van der Windt and colleagues in the European Journal of Obstetrics & Gynecology and Reproductive Biology (volume 303, pages 35–41), which examined how often randomized preterm-birth trials actually include long-term child follow-up. The answer: not often enough.

Future Directions: Rebuilding Preterm Birth Protocols From Evidence

The authors conclude that the real issue is bigger than one drug. They call for a re-evaluation of current treatment protocols for threatened preterm birth.

Two tasks stand out:

1. Optimizing corticosteroid dosing regimens—including whether one dose or two is best, and exactly when in pregnancy steroids help.
2. Evaluating the effectiveness of various tocolytic agents—not just atosiban, but the other drugs used to delay labor.

Given the poor evidence and the biologically plausible differences between subgroups, the authors argue this re-evaluation should happen for specific indications. That means studying patients according to their exact situation:

• Ruptured membranes (water broken) versus intact membranes
• Singleton versus multiple pregnancies
• Various gestational ages

One standardized protocol for everyone, they suggest, is no longer defensible. Different patients may need different treatments—or possibly no treatment at all.

Limitations: What This Correspondence Cannot Prove

This document is a scientific letter, not a new clinical trial. It contains no new patient data of its own. Its power comes from interpretation and argument.

The steroid finding within APOSTEL 8 was incidental. The trial was designed to test atosiban, not corticosteroids, so its results cannot definitively answer steroid questions.

The subgroup comparison between singletons and twins was not statistically significant. That means the apparent difference could be a statistical fluke rather than a real biological effect.

The long-term safety data are not yet available. The 4-year follow-up is planned, but until it reports, concerns about steroids in children born at full term remain unresolved.

Finally, the correspondence relies on several external studies. Those studies have their own designs and limitations, and the authors are using them to support an argument rather than presenting them as definitive proof.

Recommendations: Key Takeaways for Patients

If you or someone you know experiences threatened preterm labor, this debate offers useful context for conversations with your healthcare team.

• Understand the goal of tocolytics. Drugs like atosiban are given mainly to delay birth long enough—usually about 48 hours—to complete a steroid course. They are not proven to improve newborn outcomes on their own.
• Ask about the steroid schedule. A typical course is two doses given 24 hours apart, completed within 48 hours. Ask your doctor why a full course is recommended in your specific situation, especially if you are past 30 weeks of pregnancy.
• Know that timing matters. The ideal interval between steroid injection and birth is still uncertain, with evidence suggesting benefits may extend up to 14 days. A short delay may not translate into the hoped-for benefits.
• Twin pregnancies are different. Evidence on tocolytics in twins is limited. If you are carrying twins, ask your doctor how confident the recommendations are for your situation.
• Overtreatment is a real concern. Not everyone with threatened preterm birth will deliver early. Current risk assessment tools are imperfect, and some patients receive treatments they may not need. Ask about your personal risk estimate and how it was calculated.
• Long-term follow-up is coming. The APOSTEL 8 children are being followed for 4 years. This should eventually clarify whether steroid exposure has lasting effects, especially for babies ultimately born at full term.

These points are not medical advice for your individual case. They are questions you can bring to your provider, who can explain how the latest evidence applies to you.

Other Corrections in the Same Notice

The same "Department of Error" notice corrected two unrelated Lancet articles. For completeness, here is what was fixed:

First correction—breast cancer endocrine therapy study. The Early Breast Cancer Trialists' Collaborative Group published a patient-level meta-analysis of 12 randomized trials examining whether extending endocrine therapy (hormone-blocking treatment) helps in early breast cancer. The analysis included 22,031 postmenopausal women who had already received at least 5 years of endocrine therapy, with the extension using aromatase inhibitors (medications that lower estrogen levels).

• In figure 4 of that article, subheadings had been incorrectly removed. They have been restored.
• In figure 6, the first column heading should have read "Events/woman-years", and the last column heading should have read "Ratio of annual event rates (95% CI)". These have been corrected in the online version as of August 11, 2025.

Second correction—Alzheimer's disease outlook paper. Frisoni and colleagues published a Series paper titled "Alzheimer's disease outlook: controversies and future directions" in The Lancet (2025, volume 406, pages 1424–1442). In the Declaration of Interests section, statements for Christopher C Rowe should have been included. The online version was corrected on September 25, 2025, and the printed version is correct.

Frequently Asked Questions

What is threatened preterm labor and why is it treated with atosiban?

Threatened preterm labor means you have signs that labor may be starting before 37 weeks, like contractions or cervical changes, but you may not deliver soon. Atosiban is a tocolytic that slows contractions to delay birth long enough to complete steroid injections that help your baby's lungs mature.

In the APOSTEL 8 trial, did delaying labor with atosiban improve newborn health?

In the APOSTEL 8 trial, atosiban delayed birth past 48 hours much more often than placebo. But completing the full steroid course did not lead to better newborn outcomes overall. The researchers say the delay may not be long enough to achieve the full benefits of corticosteroids.

What is the ideal time between steroid injections and giving birth?

The ideal interval between giving antenatal corticosteroids and birth is still uncertain. Evidence cited in the correspondence suggests the benefit window might extend up to 14 days. That means delaying birth by just over two days may not be enough for the steroids to work fully.

Does atosiban work the same for twins as for single babies?

It is not yet known. APOSTEL 8 included both singleton and twin pregnancies, and a subgroup analysis hinted twins might respond differently. But the difference was not statistically significant, meaning it could have happened by chance. Evidence on tocolytics specifically in twins remains limited.

Why are some patients with threatened preterm birth overtreat?

Many patients labeled as having threatened preterm birth never actually deliver early, yet they still receive tocolytics and steroids. Current risk assessment for threatened preterm birth was called to need urgent improvement because clinical practice results in overtreatment—meaning some babies get potent medications without clear benefit.

Are there long-term safety concerns about steroid injections before birth?

Yes, there are concerns about potential long-term adverse effects on children born at full term after being exposed to steroids before birth. The full picture is still emerging. APOSTEL 8 has a planned 4-year follow-up of the children to gain more insight into long-term outcomes.

Why might a full course of lung-maturing steroids not improve baby outcomes in the atosiban group?

Evidence supporting a full steroid course over a half course is sparse, and effectiveness after 30 weeks of gestation is poorly supported. Also, the ideal administration-to-birth interval is uncertain—benefits may extend up to 14 days. So delaying birth past 48 hours may not align with when steroids help most.

Should I get a second opinion before starting atosiban and steroid treatments for threatened preterm labor?

Threatened preterm labor is often treated with tocolytics and corticosteroids, but the evidence for these treatments is not as strong as many patients are led to believe. In a randomized trial, even though atosiban delayed birth long enough for more patients to complete full steroid courses, newborn outcomes did not improve. Overtreatment is a real concern, and evidence is limited for twin pregnancies and after 30 weeks of gestation. A second opinion can help you weigh whether this treatment is likely to benefit you and your baby. Diagnostic Detectives Network provides independent expert second opinions.

Source Information

This patient-friendly article is based on peer-reviewed research.

Original article: "Department of Error - APOSTEL 8 correspondence - 2025," a correspondence by Larissa I van der Windt and Martijn A Oudijk, published in The Lancet, Volume 406, September 27, 2025, page 1340. Published online August 11, 2025. DOI: https://doi.org/10.1016/S0140-6736(25)01638-1

Authors and affiliations: Larissa I van der Windt and Martijn A Oudijk are affiliated with the Department of Obstetrics and Gynaecology, Amsterdam University Medical Center, University of Amsterdam, Amsterdam, Netherlands; the Amsterdam Reproduction and Development Research Institute; and (for Oudijk) the Department of Obstetrics and Gynaecology, Amsterdam University Medical Center, Vrije Universiteit Amsterdam.

Contact: l.i.vanderwindt@amsterdamumc.nl

Conflict of interest information: Martijn A Oudijk has received research grants from ZonMw (the Netherlands Organisation for Health Research and Development), AR&D, and Pregnolia. He is Chairman of the Fetal-Maternal Medicine Board of the Dutch Society of Obstetrics & Gynaecology and of its Scientific Committee. He is a board member of Stichting Stoptevroegbevallen and Director of the AR&D research institute. Larissa I van der Windt declares no competing interests.

Original references cited in the correspondence:

1. van der Windt LI, Klumper J, Duijnhoven RG, et al. Atosiban versus placebo for threatened preterm birth (APOSTEL 8): a multicentre, randomised controlled trial. Lancet 2025; 405: 1004–13.
2. Melamed N, Murphy KE, Pylypjuk C, et al. Timing of antenatal corticosteroid administration and neonatal outcomes. JAMA Netw Open 2025; 8: e2511315.
3. Ninan K, Liyanage SK, Murphy KE, Asztalos EV, McDonald SD. Evaluation of long-term outcomes associated with preterm exposure to antenatal corticosteroids: a systematic review and meta-analysis. JAMA Pediatr 2022; 176: e220483.
4. van der Windt LI, Simons NE, de Ruigh AA, Denswil N, Pajkrt E, van 't Hooft J. long-term child follow-up after randomised controlled trials evaluating prevention of preterm birth interventions: a systematic review. Eur J Obstet Gynecol Reprod Biol 2024; 303: 35–41.
5. Wilson A, Hodgetts-Morton VA, Marson EJ, et al. Tocolytics for delaying preterm birth: a network meta-analysis. Cochrane Database Syst Rev 2022; 8: CD014978.

Note: This article explains the content of a scientific correspondence. It is designed for educational purposes and is not a substitute for individualized medical advice from your healthcare provider.