Vaginal Progesterone Supplementation in the Management of Preterm Labor: A Randomized Controlled Trial

The primary objective in this study was to evaluate the effects of vaginal progesterone supplementation for the prolongation of the latency period in preterm labor. The secondary objectives were to evaluate gestational age at delivery, rates of preterm birth less than 34 and 37 weeks, obstetric outcomes, maternal compliance with medication use, and side effects. A randomized controlled, unblinded trial was performed. Ninety women with preterm labor occurring at 24 to 34 weeks were either randomized to a vaginal progesterone group (44 women) receiving tocolytic and antenatal corticosteroids treatment combined with vaginal micronized progesterone (400 mg everyday) or to the no-progesterone group (46 women) receiving tocolytic and antenatal corticosteroids treatment only. Latency periods were more prolonged in the vaginal progesterone group than in the no-progesterone group (32.8 ± 18.7 vs. 25.8 ± 22.7 days, p = 0.045). Gestational age at delivery in the vaginal progesterone group was also higher than in the no-progesterone group (37 vs. 35 weeks, p = 0.027). There were significant reduction rates of preterm birth less than 34 weeks (13.6% vs. 39.1%, p = 0.012), low birth weight (29.5% vs. 50%, p = 0.048), neonatal respiratory distress syndrome (13.6% vs. 37%, p = 0.021), and neonatal intensive care unit admission (6.8% vs. 28.3%, p = 0.017). Combined treatment with vaginal progesterone 400 mg could prolong the latency period in preterm labor when compared with no progesterone.


Introduction
Preterm birth is one of the more common obstetric complications. The rate of preterm birth ranges from 5 to 18% (Blencowe et al., 2012). Preterm birth is diagnosed when delivery occurs between 20 and 36 +6 weeks of pregnancy (ACOG, 2016). It causes neonatal mortality and is the most common reason for antenatal hospitalization, especially when preterm birth occurs at a gestational age (GA) less than 34 weeks. Preterm birth causes neonatal prematurity-related morbidities and neonatal death (ACOG, 2016). Preterm labor is diagnosed as spontaneous labor occurring at GAs less than 37 weeks, and which generally leads to preterm birth. Thus, the proper management of preterm labor can decrease cases of preterm birth. Proper treatment during preterm labor includes tocolytic drug and antenatal corticosteroids. A tocolytic drug is a short-term medication used in preterm labor for two days. The two-day time provided allows for efforts to improve fetal lung maturity to be enacted by corticosteroid administration (Areeruk & Phupong, 2016;Theplib & Phupong, 2016).
Progesterone is a hormone with a proven ability to maintain uterine quiescence (Borna & Sahabi, 2008). It is useful for allowing a pregnancy to reach its physiologic term. At sufficient levels in the myometrium, progesterone blocks the oxytocin effect of prostaglandin F2α and α-adrenergic stimulation and therefore increases the α-adrenergic tocolytic response. Natural progesterone is free of any disturbing teratogenic, metabolic, or hemodynamic effects (da Fonseca et al., 2003;Fuchs & Fuchs, 1984).
In 2012, the American College of Obstetricians and Gynecologists recommended using prophylactic vaginal progesterone in women with a short cervical length for the prevention of preterm birth (ACOG, 2016). Various forms of progesterone such as 17α-hydroxyprogesterone caproate intramuscular injection, vaginal progesterone gel, and vaginal progesterone tablet were investigated to prevent preterm delivery in women with a singleton gestation and a prior spontaneous preterm singleton birth (ACOG, 2012). Recently, a meta-analysis of individual patient data demonstrated that vaginal progesterone decreases the risk of preterm birth in singleton gestations with a short cervix without deleterious effects on childhood neurodevelopment (Romero et al., 2018). However, there have been no recommendations for using maintenance progesterone after preterm labor management for the prevention of preterm birth. Randomized clinical studies evaluating the use of vaginal progesterone for the maintenance of tocolysis after an episode of threatened preterm labor and preterm labor have produced conflicting data. Some research found that vaginal progesterone maintenance therapy after successful parenteral tocolysis did not significantly reduce the rate of preterm birth (Arikan et al., 2011;Borna & Sahabi, 2008;de Tejada et al., 2015;Palacio et al., 2016). Some studies found a significantly longer latency period (Arikan et al., 2011;Borna & Sahabi, 2008), while one study did not find a longer latency period (de Tejada et al., 2015). Due to these conflicting results, this study sought to assess the effects of vaginal progesterone supplementation for the prolongation of the latency period in preterm labor. The secondary objectives were to assess GA at delivery, rates of preterm delivery less than 34 and 37 weeks, obstetric outcomes, maternal compliance with medication use, and side effects.

Methods
This randomized controlled, unblinded study was performed at Department of Obstetrics and Gynecology, Faculty of Medicine, University, between June 2017 and July 2018. This study was approved by the Institute Research Ethics Committee and was registered at ClinicalTrials. gov (Trial registration: ClinicalTrials.gov, Clinical trials registration no. NCT03202836. Registered 29 June 2017, https:// clini caltr ials. gov/ ct2/ show/ NCT03 202836). All participants gave written informed consent. The methods were performed in accordance with approved standard guidelines adhering to the Declaration of Helsinki.
Singleton pregnant women aged 18 to 45 years who presented to the labor room with preterm labor with intact membranes at 24 to 34 weeks of pregnancy were invited to participate. Preterm labor was diagnosed as having a regular uterine contraction combined with changes in cervical dilatation or effacement (ACOG, 2016). Exclusion criteria included pregnant women with contraindications to tocolytic drugs, maternal or fetal conditions that required immediate delivery, cervical dilatation of 5 cm or more, "incompatibility with-life" fetal anomalies, fetal death, and allergy to progesterone.
Eligible women were enrolled after the study was approved. Pelvic examination was performed to assess cervical dilatation and effacement. A tocometer was used to measure uterine contractions. Cervical effacement, dilatation, and uterine contraction data were obtained. Baseline ultrasonography was performed to confirm GA and to obtain the estimated fetal weight. All participants received oral nifedipine, oral indomethacin, or intravenous terbutaline for 48 h of tocolytic and received a total of four doses of dexamethasone (6 mg intramuscularly every 12 h). The dosage of nifedipine first involved loading 10-mg doses every 30 min for two doses, which then continued with 10 to 20 mg every 6 h. The dosage of indomethacin was started a loading 50-mg dose of indomethacin and then continued with 25 to 50 mg every six hours. The dosage of terbutaline was 5 to 10 μg/min administered intravenously.
Prior to the study, the co-investigator generated the allocation sequence. This co-investigator had no contact with participants. The opaque envelopes which contained a code for the vaginal progesterone group or no-progesterone group were sequenced in numerical order to ensure randomization. Participants were randomized into two groups: a vaginal progesterone group or no-progesterone group. A block-of-four technique by random-number table was used for randomization. Nurses at the labor room enrolled, selected a sequentially numbered opaque envelope when a study woman met the inclusion criteria, and allocated women to their respective groups.
In addition, other opaque envelopes containing 28 or 14 capsules of micronized progesterone were labeled. These opaque envelopes were used in only the vaginal progesterone group. Drugs were prepared by a pharmacist who was not involved in the study at a point prior to the study. Fourteen-capsule and 28-capsule opaque envelopes were used for women with one-week and two-week follow-up periods, respectively. Micronized progesterone (Utrogestan®200 mg/ capsule; Besins Healthcare) was assigned to the vaginal progesterone group, while the no-progesterone group did not receive this medication. Drug use was initiated at the beginning of treatment with tocolysis. The drug dosage was two capsules applied to the vagina daily. Treatment was continued until 36 +6 weeks or until delivery. Women were admitted until uterine contractions stopped. The followup protocol from a previous study was used (Areeruk & Phupong, 2016). Women recorded their uterine contractions and were followed-up with two-week and one-week intervals until delivery. If the GA was less than 36 weeks, women were followed-up with a two-week interval. If the GA ≥ 36 weeks, women were followed-up with a one-week interval until delivery. Women in the vaginal progesterone group had to return their opaque envelopes to review their compliance at each follow-up time and at delivery. Good compliance meant that the women used all prescribed study medications.
The primary outcome was to evaluate the latency period (the time from the onset of preterm labor until birth). The secondary outcomes were to evaluate GA at delivery, rates of preterm delivery less than 34 weeks and 37 weeks, obstetric outcomes, maternal compliance with medication use, and side effects. Low birth weight is defined as birth weight less than 2500 g regardless of gestational age (WHO, 2010).
The sample size was calculated from the latency period from a previous study (Arikan et al., 2011). The means ± standard deviations of the latency period were 32.1 ± 17.8 days and 21.2 ± 16.3 days in the vaginal progesterone group and control group, respectively. At least 39 participants in each group were needed to detect a statistical difference (β = 0.2, α = 0.05). To allow for a loss to followup rate of 15%, a total of 90 participants were needed in the study.

Statistical Analysis
Statistical analysis was performed by SPSS version 22 (IBM Corp., Armonk, NY, USA). An independent t-test and analysis of covariance were used for comparing continuous data. The chi-squared test and Fisher exact test were used for comparing categorical data. The Mann-Whitney U test was used for comparing nonparametric data. Kaplan Meier analysis was used to assess the latency periods by the GA at enrollment. A p value less than 0.05 was deemed statistically significant. An intent-to-treat (ITT) analysis was used for analysis purposes in this trial.

Results
Ninety women were included into this study. All participants were randomly allocated to the two groups, with 44 receiving vaginal progesterone and 46 receiving no progesterone (Fig. 1). One woman in the no-progesterone group was lost to follow-up. However, 90 women were included in the ITT analysis. Regarding demographic characteristics (i.e., age, gravidity, parity, GA, history of preterm birth, body mass index, cervical dilatation, blood pressure, type of tocolysis, and dexamethasone administration), there were no significant differences between the groups (Table 1). Various types of tocolysis (oral nifedipine, oral indomethacin, or intravenous terbutaline) were used for short term (48 h). Even terbutaline was used much more in non-progesterone group, there were no significant differences between the groups.
Latency periods were prolonged in the vaginal progesterone group more than in the no-progesterone group (32.8 ± 18.7 days vs. 25.8 ± 22.7 days, p = 0.045)( Table 2). Further, median of GA at delivery in the vaginal progesterone group was higher than no-progesterone group (37 weeks vs. 35 weeks, p = 0.027). There was also a significant reduction rate in preterm delivery occurring less than 34 weeks (13.6% vs. 39.1%, p = 0.012). Conversely, there were no differences in rate of preterm delivery less than 37 weeks, mode of delivery, pregnancy complications, re-admission rate, and length of hospital stay (Table 2). When Kaplan Meier analysis was used to assess the latency periods by GA at enrollment, latency periods were longer in the vaginal progesterone group than in the no-progesterone group at GA 30-34 weeks at enrollment (log rank = 0.048) (Fig. 2a). But latency periods were not different between group at GA 24-29 +6 weeks at enrollment (log rank = 0.956) (Fig. 2b).
Compliance in the vaginal progesterone group was good [37/44 (84%)]. Side effects occurred in 14 cases (31.8%), including six cases of vaginal discharge (13.6%), four cases of vaginal itching (9.1%), three cases of dizziness (6.8%) and one case of nausea/vomiting (2.3%). Table 3 shows neonatal outcomes. Low birth weight (LBW), respiratory distress syndrome (RDS), and admission to the neonatal intensive care unit (NICU) were significantly less prominent in the vaginal progesterone group than in the no-progesterone group. There were no differences between the groups regarding birth weight, Apgar scores of less than 7 at one and five minutes, bronchopulmonary dysplasia (BPD), intraventricular hemorrhage (IVH), necrotizing enterocolitis (NEC), sepsis, jaundice due to prematurity, days of neonatal hospitalization, and neonatal mortality.

Discussion
This randomized controlled study evaluated the efficacy of vaginal progesterone in preterm labor treated with tocolysis and corticosteroids. This study showed that the latency period was prolonged in the vaginal progesterone group more so than in the no-progesterone group. GA at delivery in the vaginal progesterone group was also more than in the no-progesterone group. Further, there were significant reduction rates in preterm delivery less than 34 weeks, LBW, and rate of neonatal respiratory distress syndrome. Similar results between groups were found in terms of the rate of preterm delivery less than 37 weeks, obstetric outcomes, other neonatal outcomes, maternal compliance with medication use, and side effects.
In this study, the latency period was prolonged more so in the vaginal progesterone group than in the no-progesterone group (32.8 days vs. 25.8 days, p = 0.045). This finding was consistent with those of prior studies that assessed vaginal micronized progestogen in women with threatened preterm labor (Areia et al., 2013;Arikan et al., 2011;Borna & Sahabi, 2008;Sharami et al., 2010). In their treatment efforts, one study used 400 mg of vaginal progesterone (Borna & Sahabi, 2008) while some studies employed 200 mg of vaginal progesterone (Sharami et al., 2010). This present finding was in contrast with Noblot et al.' study (1991) (Noblot et al., 1991). These authors assessed the maintenance of tocolysis with oral micronized progesterone 400 mg in cases of preterm labor and found no difference in the latency period (Noblot et al., 1991). Of note, differences

Exclude (n=10)
x Decline to participate (n=10) Fig. 1 Profile of patient follow-up following randomization to either vaginal progesterone or no vaginal progesterone group 1 3 between this prior research and our study may be from the different route of medication used.
The results of this study were also different from those of previous study that assessed oral dydrogesterone 20 mg in the adjunctive management of preterm labor. There, latency periods were not different between the groups (Areeruk & Phupong, 2016). The difference between the studies may be due to the difference in type of progesterone and the route of drug administration used.
Vaginal micronized progesterone was chosen for use in the present study because the vaginal route enhanced the bioavailability and the absence of undesirable side-effects such as sleepiness, fatigue, or headache (Borna & Sahabi, 2008;Friedler et al., 1999). Micronized progesterone 400 mg was used in this study, similar to in a previous study (Borna & Sahabi, 2008).
In the present study, the rate of preterm delivery less than 34 weeks was significantly reduced in the vaginal progesterone group. This finding was consistent with that in a previous study (Bomba-Opon et al., 2012), where the authors reviewed the use of vaginal progesterone 200 mg in threatened preterm labor after tocolysis. They found that vaginal progesterone reduced preterm delivery before 34 weeks when compared with no medication (Bomba-Opon et al., 2012). However, this finding was in contrast with other results in previous studies (de Tejada et al., 2015;Palacio et al., 2016;Sharami et al., 2010), where the authors found that the rates of preterm delivery less  than 34 and 37 weeks were not significantly reduced when compared with placebo (Palacio et al., 2016). The difference in results between these studies could be attributed to the variations in the study population, dose of vaginal progesterone, and time of starting progesterone.
In present study, combined treatment with vaginal progesterone could prolong the latency period in preterm labor when compared with no progesterone. However, we cannot suppose that vaginal progesterone by itself (without corticosteroids) could also be effective. Further study with vaginal progesterone should be conducted.
The rate of preterm birth in this study was much higher in the control group. The effect size was enormous in this study (39.1% vs 13.6% for preterm birth < 34 weeks). This is typically seen in preterm labor patients, which is a common critic of the MEES trial, and may be the reason in smaller cohort, that was not found in larger cohorts.

3
A recent systemic review and meta-analysis compared vaginal progesterone, oral progesterone, 17 α-hydroxyprogesterone caproate, cerclage, and pessary for preventing preterm birth in at-risk singleton pregnancies. The study notably found that vaginal progesterone was the only intervention with a consistent level of effectiveness for preventing preterm birth in singleton at-risk pregnancies overall and in those with a previous preterm birth (Jarde et al., 2019).
Neonatal outcomes including LBW, RDS, and NICU admission were reduced in the vaginal progesterone group than in the no-progesterone group. This may be explained by the higher GA at delivery and the reduced rate of preterm delivery less than 34 weeks in the vaginal progesterone group.
The mechanisms of progesterone for prolonging pregnancy are not clearly understood, but progesterone inhibits myometrial activity by way of several mechanisms. Specifically, progesterone suppresses genes that regulate uterine contractions (i.e., genes for calcium channels, oxytocin receptors, and the gap junction protein connexin), upregulates relaxation mechanisms (i.e., the generation and action of cyclic adenosine monophosphate, and cyclic guanosine monophosphate), and acts by opposing estrogen, which increases myometrial contractility (Garfield et al., 1998). In a human study from Lucovnik et al., it was found that maintenance tocolysis with 400 mg of vaginal micronized progesterone reduced the propagation velocity of electrical signals within the myometrium (Lucovnik et al., 2018).
A strength of this study was that it was a randomized controlled study conducted to assess the efficacy of vaginal micronized progesterone for both tocolytic and the maintenance treatment of preterm labor. Vaginal micronized progesterone was chosen in the present study because the vaginal route enhances bioavailability and had no undesirable known side-effects such as sleepiness, fatigue, and headache (Borna & Sahabi, 2008;Friedler et al., 1999). Micronized progesterone 400 mg was used in this study like in a previous study (Borna & Sahabi, 2008). A limitation of this study was that it was not a placebo-controlled trial and a limitation of sample size in the vaginal progesterone group. Future randomized, placebo-controlled studies of vaginal progesterone with larger sample sizes should be performed to evaluate the value of vaginal micronized progesterone as a secondary prevention method in preterm labor.

Conclusions
Combined treatment with vaginal progesterone 400 mg in preterm labor could prolong the latency period as compared with no progesterone. It also decreased the rate of preterm delivery less than 34 weeks, LBW and RDS.