New England Journal of Medicine June 12, 2003

Prevention of Recurrent Preterm Delivery by 17 Alpha-Hydroxyprogesterone Caproate (MFMU Progesterone Trial)

Paul J. Meis, Mark Klebanoff, Elizabeth Thom, Mitchell P. Dombrowski, Baha Sibai, Atef H. Moawad et al.

Bottom Line

In pregnant women with a history of spontaneous preterm delivery, weekly intramuscular injections of 17 alpha-hydroxyprogesterone caproate (17-OHPC) significantly reduced the risk of recurrent preterm birth compared to placebo.

Key Findings

1. Treatment with 17-OHPC significantly reduced the risk of delivery before 37 weeks compared to placebo (36.3% vs. 54.9%; Relative Risk [RR] 0.66; 95% CI, 0.54-0.81).
2. 17-OHPC also significantly reduced the risk of delivery before 35 weeks (20.6% vs. 30.7%; RR 0.67; 95% CI, 0.48-0.93) and before 32 weeks (11.4% vs. 19.6%; RR 0.58; 95% CI, 0.37-0.91).
3. Infants in the 17-OHPC group had significantly lower rates of severe complications, including necrotizing enterocolitis (0% vs. 2.6%) and intraventricular hemorrhage (1.3% vs. 5.2%).

Study Design

Design
RCT
Double-Blind
Sample
463
Patients
Duration
Until delivery
Median
Setting
Multicenter, US
Population Pregnant women (15 to 20 weeks of gestation) with a documented history of at least one previous spontaneous preterm delivery
Intervention Weekly intramuscular injections of 250 mg of 17 alpha-hydroxyprogesterone caproate (17-OHPC), starting at 16-20 weeks gestation and continuing until 36 weeks or delivery
Comparator Weekly intramuscular injections of a visually matched inert placebo (castor oil)
Outcome Preterm delivery before 37 weeks of gestation

Study Limitations

The trial was stopped early at the interim analysis by the data and safety monitoring committee due to apparent efficacy, which often exaggerates the true treatment effect size.
The placebo group had an unusually high background rate of preterm birth (54.9%), which may have artificially inflated the relative benefit of the 17-OHPC intervention.
The study population was heavily enriched for a specific high-risk demographic (predominantly African American women, many with more than one prior preterm birth), potentially limiting generalizability to broader or lower-risk populations.
The trial did not utilize transvaginal ultrasound cervical length screening to stratify risk or assess baseline cervical shortening.

Clinical Significance

This landmark trial temporarily transformed obstetric practice by establishing 17-OHPC as the standard of care for preventing recurrent spontaneous preterm birth, leading to the FDA's accelerated approval of Makena. However, its clinical significance was ultimately reversed when the much larger confirmatory PROLONG trial (2020) failed to replicate these benefits, resulting in the FDA withdrawing 17-OHPC from the market in 2023.

Historical Context

Prior to 2003, preventative options for recurrent preterm birth were limited. The dramatic findings of this MFMU Network trial led to rapid endorsement by the American College of Obstetricians and Gynecologists (ACOG) and the 2011 FDA accelerated approval of Makena (17-OHPC). As a condition of accelerated approval, the FDA mandated a larger confirmatory trial (PROLONG). Published in 2020, the PROLONG trial (n=1,708) found no significant difference in preterm birth <35 weeks (11.0% with 17-OHPC vs. 11.5% with placebo). Consequently, following extensive review and public hearings, the FDA officially withdrew its approval for Makena in April 2023, representing one of the most prominent reversals in modern obstetric practice.

Guided Discussion

High-yield insights from every perspective

Med Student
Medical Student

Physiologically, how does progesterone help maintain a normal pregnancy, and based on these mechanisms, why was 17 alpha-hydroxyprogesterone caproate (17-OHPC) hypothesized to prevent recurrent spontaneous preterm birth?

Key Response

Progesterone is essential for maintaining uterine quiescence. It prevents myometrial contractility by downregulating the formation of gap junctions, decreasing the expression of oxytocin receptors, and suppressing the local inflammatory pathways that lead to early cervical ripening and labor initiation. Supplementation was hypothesized to artificially prolong this quiescent state in women prone to premature labor.

Resident
Resident

Based on the original MFMU trial, what was the specific indication for 17-OHPC, and how does this clinical indication differ from the current indication for vaginal progesterone?

Key Response

The MFMU trial targeted women with a prior history of spontaneous preterm birth (a historical risk factor). In contrast, vaginal progesterone is classically indicated for women discovered to have a short cervix on transvaginal ultrasound during the current pregnancy (an anatomical risk factor), regardless of their prior obstetrical history. Distinguishing between history-indicated and ultrasound-indicated interventions is a frequent board testing point.

Fellow
Fellow

The 2003 MFMU trial demonstrated significant efficacy for 17-OHPC, yet the much larger 2019 PROLONG trial failed to replicate these findings. What were the key demographic and baseline risk differences between these two study populations that likely explain this dramatic discrepancy?

Key Response

The MFMU trial consisted of a high-risk, predominantly US-based population with a high proportion of Black women, and an unusually high recurrent preterm birth rate in the placebo group (~54% at <37 weeks). The PROLONG trial was an international study, predominantly White, with a much lower baseline preterm birth rate in the placebo group (~11%). This highlights how varying baseline risk levels can fundamentally alter absolute risk reduction and the replicability of trial findings.

Attending
Attending

Following the FDA's 2023 withdrawal of Makena (17-OHPC) due to the PROLONG trial results showing lack of efficacy, how should you counsel a pregnant patient with a history of spontaneous preterm birth who previously received and 'benefited' from 17-OHPC in a successful prior pregnancy?

Key Response

Counseling requires validating her previous experience while explaining that larger, newer datasets indicate the drug does not actually reduce neonatal morbidity or recurrent preterm birth, suggesting her prior success was likely the natural course of that pregnancy. Counseling should then pivot to current evidence-based management, prioritizing serial transvaginal cervical length screening and offering a cerclage or vaginal progesterone if cervical shortening occurs.

Scholarly Review

Critical appraisal through the lens of expert reviewers and guideline development

PhD
PhD

The MFMU trial was stopped early by the Data and Safety Monitoring Board (DSMB) due to apparent efficacy. How does early termination for benefit introduce methodological bias, and how might this 'winner's curse' have impacted the perception of 17-OHPC's efficacy for nearly two decades?

Key Response

Stopping a trial early for benefit often captures a 'random high' in the data, leading to an overestimation of the true treatment effect size (the 'winner's curse'). This inflated effect size misguides future power calculations and establishes a false, overly optimistic sense of robust efficacy that frequently fails to hold up in larger, fully completed replication studies.

Journal Editor
Journal Editor

A critical reviewer of the 2003 MFMU trial might flag the preterm birth rate of 54.9% (at <37 weeks) in the placebo group. Why is this specific statistic a major threat to the internal validity and interpretability of the relative risk reduction reported?

Key Response

A 54.9% recurrence rate is remarkably higher than historical cohorts for recurrent preterm birth, which typically hover around 20-30%. If a control group experiences an uncharacteristically high rate of the primary outcome (due to chance, unmeasured confounding, or selection bias), the intervention will artificially appear highly protective. Editors must scrutinize whether the control arm represents the true standard natural history.

Guideline Committee
Guideline Committee

Given the trajectory from the 2003 MFMU trial to the 2019 PROLONG trial and the 2023 FDA withdrawal of 17-OHPC, how have ACOG and SMFM guidelines been forced to shift, and what does this illustrate about the process of updating clinical guidelines based on post-market confirmatory trials?

Key Response

Historically, guidelines provided a strong recommendation for 17-OHPC based on the MFMU trial. Following PROLONG and the FDA withdrawal, ACOG and SMFM updated their guidance to state that 17-OHPC is no longer recommended for preterm birth prevention. This illustrates the critical importance of 'living' guidelines and Level A evidence updates that can reverse long-standing clinical dogma when FDA-mandated post-market phase 4 trials fail to demonstrate clinical benefit.

Clinical Landscape

Noteworthy Related Trials

2011

PREGNANT Trial

n = 458 · Ultrasound Obstet Gynecol

Tested

Vaginal progesterone gel (90 mg daily)

Population

Asymptomatic singleton pregnancies with a short cervix (10-20 mm)

Comparator

Placebo gel

Endpoint

Preterm birth <33 weeks of gestation

Key result: Vaginal progesterone significantly reduced the rate of preterm birth before 33 weeks and decreased neonatal morbidity.
2016

OPPTIMUM Trial

n = 1,228 · Lancet

Tested

Vaginal progesterone (200 mg daily)

Population

Singleton pregnancies at high risk of preterm birth

Comparator

Placebo

Endpoint

Birth <34 weeks, neonatal composite morbidity/death, and childhood cognitive score

Key result: Vaginal progesterone did not significantly reduce the risk of preterm birth or improve neonatal outcomes in this mixed high-risk cohort.
2019

PROLONG Trial

n = 1,708 · Am J Perinatol

Tested

17-alpha-hydroxyprogesterone caproate (17-OHPC)

Population

Singleton pregnancies with a history of spontaneous preterm birth

Comparator

Placebo

Endpoint

Preterm birth <35 weeks and composite neonatal morbidity/mortality

Key result: 17-OHPC did not significantly reduce recurrent preterm birth or improve neonatal outcomes compared to placebo.

Tailored to your role

Want this tailored to you?

Add your specialty or training stage to get role-specific takeaways and more questions.

Personalize this analysis