Glossary

Pharmacodynamics and their implications for treatment

The mechanism of action for supplemental progestogens to improve pregnancy outcome likely relies on increased interaction between progesterone receptors and their ligands. Presumably, the enhanced receptor-ligand interaction alters ≥1 hormone-mediated physiologic properties aimed at meeting the dynamic functional demands placed on tissues of the reproductive tract during pregnancy. Each tissue of the reproductive tract, the chorioamniotic membranes, and the fetus express progesterone receptors with potential physiologic activities. The potential to augment cellular and tissue functions that are mediated by progesterone receptors beyond that achievable by the hormone that is produced from the preterm placenta alone has been termed the progestogen hypothesis .

If increasing the bioavailability of progesterone for its receptors within the reproductive tract is the therapeutic target, then this goal may be realized through supplementation that increases concentration within these target tissues or by reduction of progesterone degradation. Therefore, a potential alternative site of action for progestogens is within the liver. Caritis et al reported a linear, highly significant positive correlation between serum 17-OHPC concentration and serum progesterone concentration ( R ² = 0.46; P < .0001). An association between 17-OHPC exposure and an increased serum progesterone concentration has also been observed in 2 animal models, 1 of which was a primate model (both P < .01). Furthermore, 17-OHPC and progesterone have been shown to competitively interact with the cytochrome P450 3A4 enzyme (CYP 3A4) in human liver microsome preparations. Of note, supplemental progestogens do not act to increase the placental production of progesterone or cross-react with other steroid hormone receptors. Therefore, data support 2 potential sites of action for progestogens to enhance the progesterone receptor-ligand interaction; however, each strategy may have different capabilities to alter progesterone actions within the reproductive tract.

Both progesterone and 17-OHPC have been shown to alter progesterone receptor and cellular activity, but the relative binding affinity of 17-OHPC for nuclear progesterone receptors A and B is only 26-30% that of natural progesterone. This lower binding affinity raises the question whether this synthetic drug can act with equal efficacy as natural progesterone to influence receptor-mediated activities directly within the reproductive tract. In addition to the pharmacologic properties of these drugs and their site of action, other factors influence treatment response to progestogens. The population that is treated is the most important consideration because a variation in response to these agents has been demonstrated in different populations. Furthermore, within populations, individual patient characteristics appear to alter the effectiveness of this treatment strategy. In support of the latter construct, Manuck et al demonstrated a variable response to 17-OHPC exposure, based on the progesterone receptor genotype.

In an informative experiment in a pregnant murine model, Nold et al assessed differences in gene transcription after supplemental progesterone or 17-OHPC exposure in select pathways that are implicated in the pathophysiology of preterm delivery. The effects of 17-OHPC and progesterone in the cervix and myometrium were assessed. These investigators found progesterone, but not 17-OHPC, had a treatment effect that was localized to the cervix. Specifically, defensin-1, an antimicrobial peptide, was significantly up-regulated by supplemental progesterone exposure. These data suggest that a potential relationship exists among mucosal integrity, inflammation, cervical remodeling, spontaneous preterm labor, and this treatment.

A detailed series of experiments that tested the immune response to supplemental progestogens was performed recently by Furcron et al in a pregnant murine model. These investigators also observed that vaginal progesterone, but not 17-OHPC, was associated with beneficial effects. The immune response to natural progesterone alone resulted in a significant increase in the proportion of decidual CD4+ Tregs, a reduction in the proportion of macrophages in decidual tissues, and a reduction of active MMP-9+ cells in the cervix. Furthermore, vaginal progesterone was shown to protect against endotoxin-induced preterm birth (effect size, 50%; P = .0008). Although additional study of supplemental progestogens is needed, current evidence suggests that the alteration in the immune response is an important mechanism of action for these agents.

A paucity of experimental data is available to describe the effects of supplemental progestogens on human tissues at preterm gestational ages. The treatment response to a dose of supplemental progestogens may differ throughout gestation because placental production of the hormone varies so remarkably based on gestational age. A study by Ruddock et al incubated myometrial strips that were obtained from term cesarean deliveries with different progestogens. Progesterone treatment was found to induce relaxation of the smooth muscle; 17-OHPC exposure stimulated myometrial activity. Another study from term cesarean deliveries by Kumar et al noted that exposure to progesterone significantly reduced membrane weakening.

In summary, the experimental observations in human tissue and animal models demonstrate that (1) sites of action within the reproductive tract and liver are possible after exposure to these distinct progestogens, (2) the cervix or decidua are likely primary targets to prevent preterm birth by these agents, but other activities are possible, and (3) natural progesterone may be the superior progestogen to alter gene transcription and cellular physiologies to vary the immune response to prevent spontaneous preterm birth.

Clinical trial data that has assessed pharmacodynamic responses to supplemental progestogens has focused on cervical length measurement. A planned secondary analysis of the largest trial to date measured cervical length at enrollment and at 28 weeks gestation in asymptomatic singletons. The difference in measurement at these time-points was significantly smaller, and the cervical length at 28 weeks gestation was significantly longer in women who were treated with progesterone. A slower rate of cervical change was also demonstrated in a randomized trial testing a higher dose of 17-OHPC (682 mg/wk) in a symptomatic population with an episode of preterm labor. However, 2 retrospective studies that evaluated prophylactic treatment with 250 mg/wk of 17-OHPC did not identify a treatment effect on cervical length. Regarding myometrial activity, a large observational study demonstrated a significant increase in contraction frequency after exposure to 17-OHPC, whereas natural progesterone exposure has been shown to significantly decrease contraction frequency. Therefore, clinical data also suggest that these 2 compounds should not be considered equivalent regarding their actions on the reproductive tract functions.

Pharmacodynamics and their implications for treatment

The mechanism of action for supplemental progestogens to improve pregnancy outcome likely relies on increased interaction between progesterone receptors and their ligands. Presumably, the enhanced receptor-ligand interaction alters ≥1 hormone-mediated physiologic properties aimed at meeting the dynamic functional demands placed on tissues of the reproductive tract during pregnancy. Each tissue of the reproductive tract, the chorioamniotic membranes, and the fetus express progesterone receptors with potential physiologic activities. The potential to augment cellular and tissue functions that are mediated by progesterone receptors beyond that achievable by the hormone that is produced from the preterm placenta alone has been termed the progestogen hypothesis .

If increasing the bioavailability of progesterone for its receptors within the reproductive tract is the therapeutic target, then this goal may be realized through supplementation that increases concentration within these target tissues or by reduction of progesterone degradation. Therefore, a potential alternative site of action for progestogens is within the liver. Caritis et al reported a linear, highly significant positive correlation between serum 17-OHPC concentration and serum progesterone concentration ( R ² = 0.46; P < .0001). An association between 17-OHPC exposure and an increased serum progesterone concentration has also been observed in 2 animal models, 1 of which was a primate model (both P < .01). Furthermore, 17-OHPC and progesterone have been shown to competitively interact with the cytochrome P450 3A4 enzyme (CYP 3A4) in human liver microsome preparations. Of note, supplemental progestogens do not act to increase the placental production of progesterone or cross-react with other steroid hormone receptors. Therefore, data support 2 potential sites of action for progestogens to enhance the progesterone receptor-ligand interaction; however, each strategy may have different capabilities to alter progesterone actions within the reproductive tract.

Both progesterone and 17-OHPC have been shown to alter progesterone receptor and cellular activity, but the relative binding affinity of 17-OHPC for nuclear progesterone receptors A and B is only 26-30% that of natural progesterone. This lower binding affinity raises the question whether this synthetic drug can act with equal efficacy as natural progesterone to influence receptor-mediated activities directly within the reproductive tract. In addition to the pharmacologic properties of these drugs and their site of action, other factors influence treatment response to progestogens. The population that is treated is the most important consideration because a variation in response to these agents has been demonstrated in different populations. Furthermore, within populations, individual patient characteristics appear to alter the effectiveness of this treatment strategy. In support of the latter construct, Manuck et al demonstrated a variable response to 17-OHPC exposure, based on the progesterone receptor genotype.

In an informative experiment in a pregnant murine model, Nold et al assessed differences in gene transcription after supplemental progesterone or 17-OHPC exposure in select pathways that are implicated in the pathophysiology of preterm delivery. The effects of 17-OHPC and progesterone in the cervix and myometrium were assessed. These investigators found progesterone, but not 17-OHPC, had a treatment effect that was localized to the cervix. Specifically, defensin-1, an antimicrobial peptide, was significantly up-regulated by supplemental progesterone exposure. These data suggest that a potential relationship exists among mucosal integrity, inflammation, cervical remodeling, spontaneous preterm labor, and this treatment.

A detailed series of experiments that tested the immune response to supplemental progestogens was performed recently by Furcron et al in a pregnant murine model. These investigators also observed that vaginal progesterone, but not 17-OHPC, was associated with beneficial effects. The immune response to natural progesterone alone resulted in a significant increase in the proportion of decidual CD4+ Tregs, a reduction in the proportion of macrophages in decidual tissues, and a reduction of active MMP-9+ cells in the cervix. Furthermore, vaginal progesterone was shown to protect against endotoxin-induced preterm birth (effect size, 50%; P = .0008). Although additional study of supplemental progestogens is needed, current evidence suggests that the alteration in the immune response is an important mechanism of action for these agents.

A paucity of experimental data is available to describe the effects of supplemental progestogens on human tissues at preterm gestational ages. The treatment response to a dose of supplemental progestogens may differ throughout gestation because placental production of the hormone varies so remarkably based on gestational age. A study by Ruddock et al incubated myometrial strips that were obtained from term cesarean deliveries with different progestogens. Progesterone treatment was found to induce relaxation of the smooth muscle; 17-OHPC exposure stimulated myometrial activity. Another study from term cesarean deliveries by Kumar et al noted that exposure to progesterone significantly reduced membrane weakening.

In summary, the experimental observations in human tissue and animal models demonstrate that (1) sites of action within the reproductive tract and liver are possible after exposure to these distinct progestogens, (2) the cervix or decidua are likely primary targets to prevent preterm birth by these agents, but other activities are possible, and (3) natural progesterone may be the superior progestogen to alter gene transcription and cellular physiologies to vary the immune response to prevent spontaneous preterm birth.

Clinical trial data that has assessed pharmacodynamic responses to supplemental progestogens has focused on cervical length measurement. A planned secondary analysis of the largest trial to date measured cervical length at enrollment and at 28 weeks gestation in asymptomatic singletons. The difference in measurement at these time-points was significantly smaller, and the cervical length at 28 weeks gestation was significantly longer in women who were treated with progesterone. A slower rate of cervical change was also demonstrated in a randomized trial testing a higher dose of 17-OHPC (682 mg/wk) in a symptomatic population with an episode of preterm labor. However, 2 retrospective studies that evaluated prophylactic treatment with 250 mg/wk of 17-OHPC did not identify a treatment effect on cervical length. Regarding myometrial activity, a large observational study demonstrated a significant increase in contraction frequency after exposure to 17-OHPC, whereas natural progesterone exposure has been shown to significantly decrease contraction frequency. Therefore, clinical data also suggest that these 2 compounds should not be considered equivalent regarding their actions on the reproductive tract functions.

Is there evidence of superiority between progestogens?

Given its results, the phase III study by Meis et al stimulated design and execution of numerous additional large trials that tested progestogens in a variety of populations. A significant reduction in recurrent preterm birth (36.3% vs 54.9%; relative risk [RR], 0.66 [95% confidence interval (CI), 0.54–0.81]) was demonstrated with exposure to 17-OHPC in addition to fewer deliveries at <35 weeks gestation (20.6% vs 30.7%; RR, 0.67 [95% CI, 0.48–0.93]). Although this trial demonstrated efficacy, several concerns have been identified that include an imbalance between study groups, patient selection and generalizability of its conclusions because of the higher rate of preterm birth in the placebo group compared with other prospective observational studies, and the vehicle (castor oil) for this progestogen. Despite randomization, the women in the placebo group had a significantly higher mean number of previous preterm births (1.6 ± 0.9 vs 1.4 ± 0.7; P = .007) and a significantly greater percentage of these women had >1 previous preterm delivery before enrollment, (41.2% vs 27.7%; P = .004 by chi square based on data presented). A confirmatory phase III trial was required by the Food and Drug Administration (FDA) for its current conditional approval, given these concerns. This study has a planned enrollment of 1707 women and is intended to replicate the efficacy of 17-OHPC as a prophylactic treatment in women with a history of preterm birth. (The PROLONG Trial-clinicaltrials.gov/ct2/show/NCT01004029) This study will also better assess the safety of the drug in singleton gestations, given its planned sample size.

The FDA has also evaluated vaginal progesterone that is indicated for women with a sonographic short cervix. The marketing of natural progesterone indicated for this biomarker was not approved. In their analysis, the Agency publicly focused on treatment site interaction, the selection of statistical tests, and potential confounding variables. An independent review of the trial data differed from the FDA analysis and noted treatment benefit. Unfortunately, data from phase III trials by Fonseca et al and O’Brien et al was discounted during the review process, and a participant level metaanalysis by Romero et al was also inappropriately criticized. To date, the actions of the FDA regarding progestogens have resulted in approval of a synthetic hormone that is indicated for a risk factor, but the parent natural hormone remains unapproved when indicated for a well-validated biomarker. Therefore, these regulatory actions may have given the impression that 17-OHPC is the superior progestogen and that a treatment strategy based on a prophylactic approach is better. However, further assessment is needed.

The largest comparative efficacy trial to date that has tested the relative efficacy of progestogens for preterm birth prevention was performed by Maher et al. This trial enrolled women with more selective inclusion criteria and who were at greater risk for recurrent preterm birth. Inclusion criteria included (1) had a history of ≥1 mid-trimester preterm births or (2) had a cerclage suture placed in a previous pregnancy. Cerclage was indicated per their protocol for an obstetric history of ≥2 mid-trimester preterm births or cerclage placement in a previous pregnancy. Women with a short cervix <25 mm at <19 weeks gestation and those with or planning a cervical cerclage were excluded. This trial enrolled 518 women; outcomes were available for 253 participants who were administered 90 mg vaginal progesterone daily and 249 participants who underwent weekly intramuscular injections of 250 mg 17-OHPC. The compliance with each treatment was excellent. A significant reduction in preterm birth at <34 weeks gestation was observed with supplemental vaginal progesterone (16.6% vs 25.7%; RR, 0.58; 95% CI, 0.37–0.89; P = .02), and significantly lower rates of delivery at <32 and <28 weeks gestation were secondary findings. Pregnancy duration was in favor of natural progesterone by Kaplan-Meier analysis ( P = .0023). Finally, the number of neonatal intensive care unit admissions was lower with progesterone (15.4% vs 25.7; P = .006). Therefore, this study found that the efficacy of vaginal progesterone was superior to 17-OHPC in the prevention of recurrent preterm birth in a higher risk, compliant population.

One potential explanation for decreased efficacy of the synthetic drug is that 630 mg/week of natural progesterone was administered vs 250 mg/week of 17-OHPC. In support of this concern, Caritis et al found that the women in the lowest quartile of serum 17-OHPC concentration had higher rates of recurrent preterm birth. These data suggest, for singletons, that the optimal dose of 17-OHPC has not been identified and may vary based on factors such as body mass index, coexisting drug exposures, or innate differences in hepatic metabolism that may alter the serum concentration of this synthetic progestogen and/or natural progesterone. If monitoring serum 17-OHPC concentration is ultimately deemed necessary to optimize treatment, then intramuscular dosing of this synthetic drug will be more difficult and costly. The vaginal route of drug delivery likely results in a greater concentration of supplemental progesterone within the uterus and cervix compared to serum, if vaginal absorption during pregnancy is similar to nonpregnant women. A “first uterine pass effect” has been documented in nonpregnant women.

The comparative efficacy study by Maher et al also addresses a potential misperception that these individual progestogens are exclusively efficacious in particular populations that are at risk. Because both of these progestogens are members of the same class of drugs, most likely ultimately targeting changes in cellular and tissue physiology that are mediated via similar receptors, natural progesterone has the potential to alter outcomes in patients who are at risk like 17-OHPC. Clinical trials that have applied a prophylactic strategy in women with a history of preterm birth likely have enrolled heterogeneous groups of women with differing pathways that led to their birth histories which may explain some variation in trial findings. The negative progesterone trial by O’Brien et al was not powered to assess efficacy in the subpopulation of women with a history of mid-trimester preterm birth and selection bias that was related to cervical length measurement at enrollment likely reduced the number of women who could respond to the intervention.

Indications for vaginal progesterone and 17-OHPC

A history of preterm birth, a sonographic short cervix, or both clinical problems have served to provide indication for this intervention. As noted previously, women at highest risk with a history of mid-trimester preterm birth should undergo prophylaxis with a progestogen. However, women with a history of later spontaneous preterm birth in the third trimester may not require this intervention. O’Brien et al and DeFranco et al evaluated treatment response in the largest randomized trial performed to date in singletons. In this study, prevention of recurrent preterm birth was not demonstrated with progesterone treatment based on historic factors alone, but a treatment response was identified when cervical length was used to stratify the population. This investigation identified a therapeutic effect to prolong gestation in women with a history of preterm birth who had a cervical length of ≤30 mm (n = 116; P = .04). Therefore, women with a history of spontaneous preterm birth in the third trimester who undergo cervical surveillance and have a cervical length >30 mm may not benefit because a positive treatment response has not been replicated for this subpopulation.

The importance of cervical length in defining the indication for treatment in asymptomatic patients has been verified by 2 other phase III trials that tested a universal screening strategy. Fonseca et al and Hassan et al demonstrated that treatment indicated for a sonographic short cervix can reduce the rate of early preterm birth in women who undergo a universal screening strategy by transvaginal ultrasound scanning. Romero et al also quantified a 42% reduction using this strategy in an individual patient-level metaanalysis. Unfortunately, 2 large trials that tested 17-OHPC in asymptomatic women with a short cervix did not identify benefit in groups with either a high-risk or low-risk profile for preterm birth. Furthermore, the study by Grobman et al did not observe a therapeutic effect in the subpopulation of low-risk women with the shortest cervical lengths, <15 mm. Rozenberg et al also did not demonstrate benefit with 17-OHPC exposure in a randomized trial of symptomatic patients with a short cervix. Therefore, vaginal progesterone appears to be the superior progestogen for women with a sonographic short cervix. Other biomarkers in concert with cervical length measurement ultimately may facilitate a better definition for treatment indication and of treatment response.

The most confident conclusion that can be made from randomized trials that have been performed to date is that prophylactic progestogen exposure in women with a multiple gestation in unselected cohorts is ineffective and potentially harmful when 17-OHPC is used as the intervention. Numerous phase III trials have all failed to demonstrate positive results for their primary endpoints. Further study of a prophylactic strategy with vaginal progesterone in multiples without risk stratification is not warranted.

Other populations with negative results from larger trials include symptomatic cohorts with preterm contractions or premature rupture of the membranes. However, 2 metaanalyses that synthesized data from smaller trials have suggested progestogens may be beneficial in symptomatic populations with preterm labor. Additional larger trials are necessary to better assess the efficacy of supplemental progesterone in women with preterm labor; treatment is not recommended in symptomatic populations until such studies are performed.

A summary of clinical investigations to date is presented in the Box . Three important constructs regarding an indication for treatment are derived from the present data: (1) defining an indication for treatment primarily based on a biomarker in the ongoing pregnancy is likely optimal to basing therapy solely on obstetric history for the majority of patients, except for those at highest risk (a previous mid-trimester spontaneous preterm birth), (2) vaginal natural progesterone appears superior to 17-OHPC for efficacy in asymptomatic patients with a short cervix, and (3) women with a multiple gestation and symptomatic patients do not respond to this therapeutic strategy like asymptomatic singleton cohorts. The strength of recommendation for the use of progestogens to prevent preterm birth is based on the Grading of Recommendations Assessment, Development, and Evaluation guidelines ( Box ), and the present review is the first to assess this treatment strategy using these suggested guidelines.

Box

Proposed Grading of Recommendations Assessment, Development, and Evaluations for supplemental progestogens and the suggested superior agent ^a

I. History indicated: prophylactic therapy in an asymptomatic, singleton pregnancy in women with a history of spontaneous preterm birth Indicated: Prior mid trimester spontaneous preterm birth — superior agent: natural progesterone. Studied agents included 17-OHPC, which is an option. Reproducible evidence is observed. Secondary analysis by Spong et al of the trial by Meis et al identified efficacy for this subpopulation. Validation for efficacy is provided from evidence for superiority of natural progesterone compared with 17-OHPC, Maher et al (assumes 17-OHPC exposure in this trial would not cause harm). Therapy should be initiated at 15-20 weeks gestation. Validation of natural progesterone being superior is needed by additional highly powered pragmatic trials. Grade 1B a Considered: Prior spontaneous preterm birth at 28-36 weeks gestation—studied agents: synthetic 17-OHPC and natural progesterone Efficacy with 17-OHPC treatment is suggested from a single phase III trial by Meis et al , but remarkable heterogeneity is present in metaanalysis (Cochrane review ). The largest trial testing vaginal progesterone did not demonstrate benefit in this population, but the prevalence of a short cervix was relatively low in this study. Objective data, particularly cervical length, likely serves to better guide therapy in this subpopulation based on data from DeFranco et al (later). The superior progestogen optimizing both efficacy and safety in this subpopulation has not been defined clearly. If treated, therapy should begin at 15-20 weeks gestation. Additional dose-ranging and safety studies are also needed. Insufficient evidence from phase III trials for strongest recommendation. Grade 2B II. Biomarker-based indication for therapy in a singleton pregnancy A. Short cervix: asymptomatic, mid trimester, 19-24 weeks gestation 1. Universal screening strategy to identify a short cervix Indicated: Cervical Length ≤20 mm—superior agent: natural progesterone. Studied agents included 17-OHPC, which should not be prescribed for this indication. Reproducible evidence for efficacy from phase III trials, (Hassan et al and Fonseca et al ). The superior agent is suggested because of negative trials/secondary analysis by several groups. Safety and efficacy in patients with the shortest of cervical lengths, <10 mm, may be the subject for additional study to define optimal intervention, particularly if fetal membranes are exposed or intrauterine infection is suspected. The efficacy of treatment at longer cervical lengths, such as 21-25 mm, also requires validation in much larger trials; however, metaanalysis by Romero et al has suggested potential benefit. Grade 1A 2. Select subpopulation: history of spontaneous preterm birth with a short cervix Considered: Cervical length ≤30 mm—studied agent: natural progesterone Based on evidence for pregnancy prolongation from a phase III trial over the continuum of cervical length by DeFranco et al and a participant-level metaanalysis by Romero et al. The utilization of this criterion may be limited to those undergoing cervical surveillance. The efficacy of treatment at cervical lengths >20 mm in this subpopulation requires validation in phase III trials. Insufficient evidence for strongest recommendation. Grade 2B B. A short cervix with symptomatic preterm labor, premature rupture of membranes, or at later gestational ages No indication at present and safety concerns exist from experimental and animal data. Further investigation reserved for carefully monitored, phase III randomized trials. A single phase III trial, Martinez de Tejada et al , demonstrated possible harm in women who are diagnosed with preterm labor exposed to progesterone. Grade 2B C. Biochemical biomarker or other sonographic finding as an indication or exclusion for therapy No indication or exclusion for treatment based on objective biochemical findings or other evidence of a potentially altered physiology has been identified and validated. III. Multiple gestations Contraindicated in unselected populations: 17-OHPC should not be further investigated or prescribed in multiples because of identified safety findings in an individual participant data metaanalysis by Schuit et al. Grade 1B Subpopulations with an objective biomarker for decidual/placental dysfunction have not been studied adequately with natural progesterone. Also, 2 individual participant level metaanalyses have suggested that a short cervix at <24 weeks gestation in asymptomatic patients may provide an indication worthy of study. Because of the presence of risk, treatment is not recommended until additional phase III trials have been performed in such candidate subpopulations that are stratified by biomarkers.

Only gold members can continue reading. Log In or Register to continue

Share this:

• Click to share on Twitter (Opens in new window)
• Click to share on Facebook (Opens in new window)

Related

Related posts:

1. Recurrent vulvovaginal candidiasis
2. Solitomab, an EpCAM/CD3 bispecific antibody construct (BiTE), is highly active against primary uterine serous papillary carcinoma cell lines in vitro
3. Neurodevelopmental outcome at 2 years in twin-twin transfusion syndrome survivors randomized for the Solomon trial
4. Association is not causation!

Stay updated, free articles. Join our Telegram channel

Join