The PROLONG Trial (Progesterone for Preterm Birth, Part II)

/

Last week, we introduced our major progesterone trial series with a discussion of the Meis trial. We finish up this week talking about the follow up trial: the PROLONG study.

The PROLONG Trial 

Methods

  • Location

    • 93 total centers across 9 countries  

  • Eligibility criteria and exclusion criteria: pretty much the same as the earlier trial 

  • Randomization and treatment 

    • Randomized 2:1 to receive 17OHP weekly, to start between 16-20w6d 

    • Placebo was benzyl benzoate, caster oil, benzyl alcohol 

  • Outcomes 

    • Measured preterm birth <35 weeks 

    • Composite neonatal morbidity and mortality index 

    • Secondary outcomes: 

      • Neonatal death, PTB <32 weeks, PTB <37 weeks and also medical indicated preterm births 

    • Primary safety outcome: fetal/early infant death 

  • Stats 

    • Unlike previous trial, wanted it to have adequate power to identify both the primary efficacy and the primary safety analyses 

    • Therefore, needed 1665 liveborn infants to detect a reduction of 35% in the rate of the composite index 

    • Needed 1707 patients to provide 98% power to detect a reduction of approximately 30% in the rate of PTB <35w0d

Results 

  • Timeline: November 12, 2009 - October 8, 2018 

  • Total of 1740 women were consented and agreed to study 

    • 1130 allocated to 17OHP, 578 allocated to placebo 

    • 390 were randomized in the US, 1317 randomized in non-US 

  • Demographics were similar between groups 

    • 88.8% vs. 87.2% Caucasion in 17OHP vs. placebo groups 

    • Prepregnancy BMI 23 

    • Almost 90% were married, living with partner compared to just about 50% in the other study 

    • Only 7-8% smoked in pregnancy compared 20-22% in other study 

  • So… similar for each arm of this study, but very different demographic compared to the Meis trial.

  • Primary Outcome 

    • Preterm birth prior to 35 weeks: 11.0% in 17OHP vs. 11.5%

      • Spontaneous was 8.4 vs 8.9%, RR 0.93, 95% CI 0.67-1.30

    • Not a primary outcome, but PTB <37 weeks: 23.1% vs 21.9% RR 1.06, 95% CI 0.88-1.28! 

      • So different from 36 vs. 50% in Meis study 

    • Neonatal outcomes: 

      • Unsurprisingly, given no diff in preterm labor, no diff in composite neonatal morbidity and mortality 

  • Secondary Outcome 

    • Preterm birth <32 weeks - also not different 

    • Other things: cerclage, preterm labor eval, tocolysis, corticosteroids, maternal GDM, preeclampsia, chorio, abruption, CS all not different 

    • Neonatal: no difference in anything, including neonatal death, RDS, NEC, IVH, NICU admission, birth weight, TTN, PDA, ROP

Conclusions

  • Study was done in consultation with the FDA as part of approval for use of 17OHP 

  • Unlike the findings in the MFMU trial, this study had much lower event rate of PTB and did not confirm treatment efficacy 

Discussion points

  • While the Meis study was conducted in the USA, the PROLONG study was mostly not, and that was because the findings from the Meis study had changed practice 

  • Most places in the US were already prescribed 17OHP and likey didn’t want to change their practice 

  • Also, the Meis showed a large predominance of black women (60%), while there were very few in the PROLONG study 

  • Finally, there was a huge discrepancy in baseline preterm labor rates between the two studies, with the rates in PROLONG about ½ has much as that of the Meis study 

  • Why castor oil? 

    • Not that it has shown to cause preterm labor, but castor oil can cause contractions, usually due to GI distress when ingested 

  • Consequences of the Studies 

  • Follow up to the study 

    • EPPPIC - Evaluating Progesterone for Preventing Preterm birth International Collaborative - meta-analysis of individual participant data from randomized controlled trials 

      • Published in the Lancet in 2021 

      • Basically: 31 trials were included with individual participant data

        • From these studies, it seems that vaginal progesterone and 17OHP both reduce birth before 34 weeks gestation in high risk singleton pregnancies 

        • Absolute risk reduction is greater for women with short cervices 

        • The data seems a little better for vaginal progesterone in consistently decreasing preterm birth <37 weeks and <34 weeks

        • For 17OHP the data just crosses the line of no effect at <34 weeks 

The Meis Trial (Progesterone for Preterm Birth, Part I)

/

One of the hottest controversies in OB in the last several years has been about the role of progesterone for treating preterm birth. We covered these trials very briefly in our episode on preterm birth prevention. Today, let’s dive deeper and start with part I - talking through the Meis Trial.

Next week, we’ll talk in depth about the PROLONG trial and some takeaways for your practice.

Background: 

  • Titles 

    • The Meis Trial - Prevention of Recurrent Preterm Delivery by 17 Alpha-Hydroxyprogesterone Caproate 

    • The PROLONG Trial - 17-OHPC to Prevent Recurrent Preterm Birth in Singleton Gestations (PROLONG Study): A multicenter, International, Randomized Double-Blind Trial 

  • Who wrote the papers and where were they published? 

    • Meis: Dr. Paul Meis!  - Done with the NICHHD as part of the MFMU 

      • Published in the NEJM! Very fancy - published 2003 

    • PROLONG: Done by Dr. Sean Blackwell (was a president of SMFM!) and a bunch of people who are active in SMFM like Dr. Cynthia Gyamfi-Bannerman, Dr. Brenna Hughes, Dr. Judette Louis 

      • Published in American Journal Perinatology in 2020; which is a good journal but… nowhere near the impact factor of NEJM lol 

  • Who funded the studies? 

    • Meis: NICHD

    • PROLONG: AMAG Pharmaceuticals (those that make Makena) 

  • Why were the studies done? 

    • The initial study (Meis) was done to try and decrease preterm delivery (leading cause of infant mortality and morbidity in the US!). There had been small trials with prophylactic treatment with progestational compounds that had shown promise, but not all of them had positive results 

      • Had been a meta-analysis restricted to 17-OHP that showed a reduction in preterm delivery 

    • The second study was done after the first study was stopped early because it was thought to be unethical to continue recruitment given the robustness of evidence (spoilers, sorry) 

      • The FDA granted conditional approval for 17-OHP for commercial use in 2011 under the provision that another well-conducted randomized controlled trial was done 

      • This was a confirmatory trial for 17-OHP, and the FDA also required a long-term infant follow-up study 

  • Question to answer: Does the use of weekly 17-OHP in pregnancy decrease the risk of preterm birth? 

The MEIS Trial

Methods

  • Done at 19 participating centers, primary USA based 

  • Eligibility criteria

    • History of spontaneous preterm delivery in previous pregnancy between 20w-36w6d 

    • Current pregnancy between 15w0d-20w3d gestation 

    • Ultrasound between 14w0d-20w6d to confirm duration of gestation and to identify any major fetal anomalies 

  • Exclusion criteria 

    • Multifetal gestation

    • Known fetal anomaly 

    • Progesterone or heparin treatment during current pregnancy 

    • Current or planned cerclage 

    • Hypertension requiring medication

    • Seizure disorder 

    • Plan to deliver somewhere other than the 19 participating center 

  • Intervention 

    • Patients that consented were randomized to either IM study drug (17 OHP) or castor oil (omg - need to talk about this after!)

      • Had to be given first dose between 16w0d-20w6d; if they did not return before 20w6d, could not participate

      •  Weekly injections until 36 weeks 

  • Outcomes 

    • Date of delivery, birth weight, neonatal course - to assess for preterm delivery!  

  • Statistics 

    • Power calculation: 

      • Estimated 37% of patients in placebo group would have a preterm delivery 

      • Sample size calculation of 500 women, with 334 in progesterone group and 166 in placebo to find a reduction of 33% in rate of preterm delivery 

    • At second interim analysis, when 463 patients had undergone randomization, outcome data was available for 351 patients

      • The boundary (p=0.015) for test of significance had been crossed for preterm delivery, and enrollment was halted 

Results

  • Timeline: 

    • Started in April 1998, but ultimately stopped in February 1999 because the FDA ordered the company that supplied to the active drug to shut down and mandated a total recall of the company’s drugs because of poor quality control and documentation 

      • 150 patients had been enrolled 

    • The study was then restarted with a new drug company. The 150 women previously with other med were not included in the study data analyses 

    • New Timeline: 

      • September 1999 - February 2002 

  • Patient Demographics 

    • 1039 eligible, 463 eligible women gave their consent  

      • 310 in progesterone group 

      • 153 in placebo group 

    • Patients were similar in mean duration of previous gestational age (that qualified them for the study), mean gestational age at time of randomization, race, marital status, BMI, smoking, education level, etc. 

      • Of note, interestingly, approximately 60% of those in the study identified as non-Hispanic black (compare to PROLONG, which has very different demographics) 

      • Prepregnancy BMI 26 

    • But, women in the placebo group had more previous preterm deliveries (1.6 vs. 1.4, p=0.007) 

    • 91.5% of women were compliant with their injections (meaning they didn’t go more than 10 days without an injection) 

    • 50% reported at least one adverse effect 

      • Most common: 34.2% reported soreness at sight of injection; 14% reported swelling, 11% itching, 6.7% bruising 

      • More women with progesterone had swelling or a lump at injection sight 

  • Primary outcome 

    • Preterm delivery

      • Data was available for 459/463 women

      • Delivery before 37 weeks was reduced, 36.3% for 17-OHP group, 54.9% for placebo group (wtf, that’s super high) RR 0.66, 95% CI 0.54-0.81 

        • Spontaneous PTL was 29.4% vs. 45.1% respectively 

      • Findings were similar for black vs nonblack women 

      • Even more interesting, significant for delivery before 35 and 32 weeks gestation

        • Delivery before 35 weeks: 20.6% vs. 30.7%, RR 0.67, 95% CI 0.48-0.93 

        • Delivery before 32 weeks: 11.4% vs 19.6%, RR 0.58, 95% CI 0.37-0.91

    • Other things were not different: hospital visit for PTL, tocolytic therapy, corticosteroid use, CS, chorio 

  • Secondary outcome 

    • Fetal and neonatal outcomes 

      • No difference in fetal death, antepartum or intrapartum 

      • There was a significant difference in birthweight

        • <2500g was 27.2% vs. 41.1%, RR 0.66 (95% CI 0.51-0.87)  

      • Also difference in supplemental oxygen use (14.9% in 17OHP vs. 23.8% in placebo) 

      • And lastly, difference in IVH (1.3% vs 5.2%) 

    • No difference in neonatal death, TTN, RDS, bronchopulmonary dysplasia, vent support, NEC, retinopathy, etc. 

Conclusions 

  • Authors concluded: treatment with 17OHP weekly from 16-20 weeks to 36 weeks of gestation reduced preterm delivery at 37, 35, and 32 weeks in patients with history of sPTB

    • But … they also did say that there was an overall very high rate of preterm delivery (50% in placebo vs. 36% for 17OHP)

    • Next week, we’ll contrast with PROLONG!

Preterm Labor and PPROM

/

Today we talk about the routine management of PPROM and PTL. We’ve prepared a little chart that we hope is handy for both teaching and learning! Be sure to also check out ACOB PB 171 and PB 188. For some primary literature, check out the BEAM trial on magnesium sulfate, the most recent Cochrane review on steroid administration, the ALPS trial for Antenatal Late Preterm Steroid administration, and the RCT demonstrating benefit to latency antibiotics in PPROM.

(c) CREOGS over Coffee, 2019

We also use the podcast to highlight a number of controversies, differing practice patterns, or areas of new and active research in these clinical topics (with help from our friends at the ObG Project!)

  • Delivery timing: A 2017 Cochrane review suggested better neonatal outcomes with expectant management of PPROM to 37 weeks, convincing enough to have the Royal College of Obstetrics and Gynecology to change their clinical practice guideline to allow expectant management to 37’0.

  • Administration of Corticosteroids: The ObG Project gives a great summary on when to administer betamethasone. In summary:

    • Between 24-34 weeks in all cases of PPROM and in PTL if delivery is expected within 7 days.

    • A single rescue course should be administered if it has been > 14 days since the last course, and delivery is again expected within the subsequent 7 days.

    • Between 34-36’6 weeks if PPROM or PTL occurs, no prior steroids have been administered, and delivery is expected within the subsequent 7 days.

  • Periviability: The management of periviable PPROM is managed very differently by institution, as resources and optimal management strategies remain to be identified. Protocols and policies should be arranged in accordance with the individual obstetrics and neonatology departments. Ideally, counseling for patients experiencing periviable PTL and PPROM should be performed in an interdisciplinary fashion.

  • Outpatient Management of PPROM: There have a few retrospective studies, the most recent of which came from a large series out of France and received some press attention, suggesting that outpatient management may be appropriate in select candidates. That said, this is definitely NOT the standard of care at this time; inpatient management of PPROM is still the standard set forth by ACOG in the absence of larger, prospective studies.