Prevention of Preterm Delivery with 17-Hydroxyprogesterone Caproate: Pharmacologic Considerations

Abstract

Despite advances in neonatal care, the burden of preterm birth remains high. Preterm birth is a multifactorial problem and strategies to identify and treat medical risk factors in early pregnancy have not been effective in reducing preterm birth rates. In a sentinel clinical trial, prophylactic therapy with 17-hydoxyprogesterone caproate (17-OHPC) reduced the risk of recurrent, spontaneous preterm birth in 34% of women¹. As a result, clinical practice changed and extensive research on 17-OHPC followed. The increasing body of evidence demonstrated a variable efficacy of the drug. This review will examine the plausibility, pharmacology, clinical efficacy and safety of 17-OHPC when used in the setting of preterm birth prevention. We will also discuss pharmacokinetic and pharmacodynamics data to highlight drug metabolism and mechanism of action; which will help clarify the variability in clinical outcomes and efficacy.

Introduction

Premature birth in the United States accounts for 35% of deaths in the first year of life at an estimated annual cost exceeding $26 billion^2,3. Efforts to identify and prevent preterm birth have led to a modest decrease in the annual rate in the United States from a peak of 12.8% in 2006 to 11.5% in 2012⁴. In addition to mortality, long-term morbidity is a significant risk of preterm birth and both are inversely related to gestational age at birth. The majority of preterm births occur in singleton pregnancies following the spontaneous early onset of the parturition process. However, certain factors are associated with an increased risk for preterm birth: multifetal gestation, shortened cervical length, genitourinary infection, smoking, and a prior preterm birth.

Clinical interventions to decrease the rate of preterm birth have focused on identifying women at increased risk for preterm birth and the use of prophylactic and therapeutic options in such subjects. Unfortunately, despite vigorous efforts at treatment of preterm labor with labor inhibiting drugs (tocolytics), intensive prenatal care, patient education and bed rest, rates of preterm birth have not decreased significantly over the past 40 years. One of the very few preventive measures to have shown some promise in randomized trials is the use of 17-OHPC, a progestational agent. In a sentinel article that drastically changed clinical practice, Meis et. al. demonstrated a 33% reduction in delivery before 37 weeks in women with a prior preterm birth, who received weekly injections of 17-OHPC¹. Treatment with 17-OHPC has proven effective in reducing the rate of preterm birth in women with a history of spontaneous preterm delivery, but it appears to be ineffective in other high-risk categories such as women with a short cervical length and those with multi-fetal gestation. This review will focus on the pharmacologic properties of 17-OHPC, which may explain the variability in outcomes noted in clinical trials.

Plausibility of progesterone for preterm birth

Progesterone is thought to act in support of gestation and to inhibit uterine activity. The concept of a “progesterone block” was advanced and championed by Csapo in the 1950’s based on his extensive and pioneering experiments in pregnant rabbits⁵. This formed the basis for the study of progesterone supplementation and the role of progesterone in the onset of labor. As such, numerous animal studies support the importance of progesterone in regulating the onset of labor^5–7. In sheep, goats, and many other mammalian species, a decrease in plasma progesterone (P) and an increase in estrogen (E) preceded the onset of labor. The ability of progesterone to maintain uterine quiescence during pregnancy has been shown in lower mammalian species. In these species, progesterone withdrawal is a necessary step in the events leading to parturition. The role of progesterone and changes in P/E ratio on the onset of labor in human beings and other primates is less well known. Although some investigators have described low progesterone concentrations or low P/E ratios in the plasma of women destined to deliver prematurely, no consistent evidence exists documenting decreases in plasma progesterone or P/E ratio prior to the onset of labor at or before term⁸. Nonetheless, some evidence exists that local changes in progesterone or the P/E ratio in the placenta, decidua, or fetal membranes may be important in the initiation of labor⁹. The progesterone withdrawal theory remains a leading hypothesis because no other mechanism for the onset of human parturition has been definitively established and because synthetic antiprogestins stimulate myometrial contractions.

In light of these findings, progesterone was theorized as an agent to prevent preterm birth. Early studies yielded conflicting results regarding its efficacy in preventing preterm birth. These studies were limited by small sample sizes and inclusion of heterogeneous patient populations. With increasing evidence, it appears that 17-OHPC therapy has condition-specific efficacy in decreasing the risk for preterm birth. A meta-analysis published by Keirse in 1990, included only studies that evaluated the effect of 17-OHPC on a range of outcomes including preterm birth. This meta-analysis found a pooled odds ratio of roughly 0.50 (95%CI: 0.30–0.85) for preterm birth among women treated with 17-OHPC¹⁰. Since then, various trials have focused on progesterone use for the prevention of preterm birth. In 463 women with singleton gestations and a history of spontaneous preterm birth, weekly 17-OHPC intramuscular injections, was associated with a decreased incidence of preterm birth (RR 0.66, 95% CI, 0.54–0.81) compared to placebo¹. In women with twin gestation, weekly 17-OHPC injections did not reduce preterm birth or adverse neonatal outcomes compared to placebo^11–13. In a multicenter randomized controlled trial, weekly 17-OHPC injections did not decrease the risk of preterm delivery in nulliparous women with a singleton gestation and a short cervical length (<30 mm)¹⁴. Furthermore, two randomized controlled trials also demonstrated a lack of benefit for 17-OHPC in triplet gestation^15,16. Based on these randomized controlled trials current guidelines recommend the use of 17-OHPC only for the prevention of preterm birth in singleton pregnancies with a history of preterm birth^17,18. As a tocolytic agent, 17-OHPC has proven unsuccessful^19–21.

Pharmacokinetic properties of 17-OHPC

Progesterone and its naturally occurring metabolite 17-hydroxyprogesterone are produced in large amounts in human pregnancy. On the other hand 17-OHPC is not a naturally occurring substance but rather is synthesized through the acetylation of 17- hydroxyprogesterone with caproic acid in the presence of toluene sulfonic acid²². The structures of progesterone, 17-OHP and 17-OHPC are depicted in Figure 1. Despite widespread usage of 17-OHPC in the 1950s through 1970s little information about the pharmacology of this agent is available. It initially gained Federal Drug Administration (FDA) approval in 1956 (NDA 10–347) and was marketed under the trade name Delalutin as a treatment for menstrual disorders (such as dysmenorrhea, pre-menstrual tension, cyclomastopathies, adenosis, and mastodynia), threatened miscarriage and uterine cancer.

Figure 1.

Chemical structure of progesterone, 17-hydroxyprogesterone and 17-OHPC

17-OHPC is a lipophilic drug and is highly protein-bound in blood. In humans, it is predominantly metabolized in the liver by the cytochrome P450 (CYP) enzymes, primarily CYP3A4 and to a lesser extent CYP3A5. Metabolites of 17-OHPC have been identified and differ from 17-hydroxyprogesterone or progesterone. The caproic acid moiety of 17-OHPC is not removed during metabolism, rather the steroid rings are primarily hydroxylated. After intramuscular administration, elimination of metabolites occurs primarily through feces (~50%) and urine (~30%). Factors that alter CYP expression and activity can affect the clearance of 17-OHPC. CYP3A activity increases with advancing gestation and genetic polymorphisms of CYP3A5 influence clearance of CYP3A5 substrates²³. Interestingly, CYP3A5 is the major CYP3A enzyme in more than half of African American adults compared to a third of Caucasians²³. This finding could explain the increased clearance noted in African American pregnant women, but fails to explain the similar plasma concentrations for all races in the same study²⁴. Not surprisingly, concomitant use of medications that are CYP3A inhibitors or inducers could lead to decrease or increase, respectively, in the clearance of 17-OHPC. At clinically relevant concentrations, 17-OHPC does not appear to have any effect on CYP1A, CYP2D6, CYP2C9 or CYP3A4, while the activity of CYP2C19 is modestly increased in primary cultures of human hepatocytes (Yang et al. AJOG In Press. Accepted March 28 2014). Dosage of CYP2C19 substrates, for example tricyclic antidepressants, proton pump inhibitors and propranolol, may have to be increased in patients on 17-OHPC. Endogenous steroids also appear to affect 17-OHPC metabolism. In human liver microsomes, a combination of steroid hormones and progesterone alone inhibited 37% and 28% of 17-OHPC metabolism, respectively²⁵. Future clinical studies should aim to substantiate the physiologic relevance of these in vitro findings.

In women with a singleton gestation receiving the recommended dose of 17-OHPC (250 mg weekly) there is a large variation in the maternal plasma concentrations of 17-OHPC (median 15.2 ng/mL, range 3.2–64.6 ng/mL) and the drug remains detectable weeks after the last injection²⁴. The half -life of 17-OHPC is reported to be 16 days (±6 days²⁴. The prolonged half-life may result from a slow release of 17-OHPC from the castor oil depot and/or the maternal adipose tissue, which may also explain the detection of the drug in maternal samples 44 days after the last intramuscular injection. Across a dosing interval, maximal plasma concentration occurred a day after drug injection (Figure 2), but samples were not taken at earlier times. The area under the plasma concentration × time curve (AUC) of 17-OHPC was higher at 31–34+6 weeks compared to 20–24+6 weeks (21.0 vs. 24.3 ng/mL, p=0.007), suggesting sustained and continuous release of 17-OHPC from multiple administration sites²⁴. In women with a singleton gestation, race did not alter drug concentrations; though the study was not powered to detect such an effect²⁴. In women with multifetal gestation 17-OHPC trough concentrations were 40% lower compared to women with a singleton gestation (9.7 vs 14.1ng/mL²⁴. Women with twins and triplets also exhibited a slightly shorter half–life (10 days²⁶. These findings were attributed to an increase in apparent clearance of 17-OHPC²⁶. Drug clearance and volume of distribution were highly variable between subjects, 24% and 49%, respectively. Fetal number was not associated with any changes in drug concentration or clearance²⁶. In both cohorts, BMI exhibited an inverse relationship with drug concentration accounting for about 20% of the variation in drug concentrations, whereas parity did not have an effect²⁴. The large variation in plasma 17-OHPC concentration may be explained by variability in the physiologic changes in pregnancy as well as maternal BMI, alterations in the metabolism of 17-OHPC by endogenous steroids, especially progesterone and competition for metabolism from other medications that are CYP3A substrates that the patient may be taking²⁵. The pharmacokinetic parameters of 17-OHPC in pregnancy are summarized in Table 1.

Figure 2. 17-OHPC concentrations during one dosing interval.

Data points represent plasma 17-OHPC concentration ± standard deviation.

Table 1. Pharmacokinetic Parameters of 17-OHPC in Pregnancy.

Absorption: Slow absorption over a prolonged time period

Distribution: Highly protein-bound in blood

High volume of distribution – extensive distribution in the body

Metabolism: Mono hydroxyl, di hydroxyl and trihydroxylated metabolites formed in liver

Excretion: Minimal unchanged drug excreted in urine and feces

	Singleton Gestation24	Twin Gestation26
CTrough (ng/mL):	14.1 ± 5.6 (10–18.1)	11.2 ± 4.6 (4.8–16.3)
CMAX (ng/mL)	22.6 ± 9.5 (15.8–27.4)	17.3 ± 6.7 (12–27)
Apparent Half-life (d)	16.2 (10.6–21.0)	10 (6–16)
TMAX (d)	1.0 ± 1.9 (1–3)	1.2 ± 0.41(1–2)
Vd/F (*103) (L)	56 (25.2–69.6)	16.9 (9.1–24.5)
Cl/F (*103) (L)	2.1 (1.5–2.7)	1.2 (0.9–1.7)

Data are mean ± standard deviation (interquartile range)

C_Trough, the minimum concentration just before the next dose; C_MAX, maximum concentration during the dosing interval; T_MAX, time to maximum concentration; Vd/F, apparent volume of distribution/bioavailability; Cl/F, clearance/bioavailability.

Pharmacodynamic properties of 17-OHPC

Although most data support the concept that progesterone plays a role in maintaining gestation in human beings, the actual mechanisms by which exogenous progesterone therapy, including 17-OHPC, may avert preterm labor and delivery are less well determined. Most studies support a likely anti-inflammatory effect and a local increase in progesterone concentrations in gestational tissues, counteracting the functional decrease in progesterone that leads to preterm birth.

The effects of 17-OHPC on the uterine cervix in pregnant women, animals, and in the context of in vitro experiments have not been studied to the same extent as those of progesterone. One study examined the effect of 17-OHPC on cervical length in patients with a history of 1 or more preterm births who were allocated to receive 17-OHPC and were compared to an untreated control group²⁷. No difference in cervical length measurements over time was observed in women who received 17-OHPC. In contrast, vaginal progesterone reduced the rate of cervical shortening in patients with a history of preterm birth or premature cervical shortening²⁸. More recently, progesterone supplementation decreased the risk of preterm delivery in asymptomatic women with a short cervix²⁹. Progesterone is also associated with a reduction in uterine contraction frequency^30,31. By comparison, ex vivo exposure to 17-OHPC is associated with no change or an increase in contraction frequency, reaffirming its lack of benefit as a tocolytic agent^30,31. Cellular preparations from cervix, decidua, and fetal membranes revealed an effect of progestins on collagen synthesis. Progesterone activity appeared to limit collagenolysis, altered production of cytokines, nitric oxide, and prostaglandins, and limited apoptosis^32–38. In contrast, progesterone antagonism resulted in impaired decidual functioning and trophoblast proliferation, and accelerated cervical ripening by collagenolysis^39–42. These differences highlight the characteristic indications and benefits for different progesterone compound and underline the need to think of each formulation distinctly.

The human progesterone receptor (PR) is a member of the steroid and thyroid reception superfamily. It is mainly located in the nucleus and exists in two isoforms (A and B⁴³. Both receptor isoforms have been described in gestational tissues including the chorion and amnion^9,43. PR isoforms likely differ in their biological roles. PR-A is thought to inhibit the transcription of progesterone receptor genes. In contrast, PR-B increases the transcription of progesterone genes and promotes myometrial quiescence⁴⁴. Thus, the responsiveness of a tissue to progesterone may depend on the concentration of circulating progesterone and the ratio of PR isoforms^45,46. In fact, it has been hypothesized that an increase in PR-A to PR-B ratio and an inhibition of PR-B activation by PR-A, may contribute to a qualitative decrease in progesterone activity and lead to the initiation of labor^47,48. Recently, studies have described an association between different single nucleotide polymorphisms (SNPs) of the PR gene promoter and preterm birth^49,50. Furthermore, Manuck et. al. described an effect of PR polymorphism on the clinical efficacy of 17-OHPC in preventing recurrent preterm birth⁵¹. In addition, a specific PR genotype was associated with an increased risk for preterm birth in Caucasian/Hispanic women who received 17-OHPC. In contrast, other genotypes were associated with a high rate of preterm birth in the placebo group and a significant reduction in recurrent preterm birth in women exposed to 17-OHPC⁵¹. Data has also emerged on the interaction between PR SNPs, BMI and race suggesting a complex genetic background of women with preterm birth^49,52.

Few studies have reported on the association between progesterone concentration and possible drug targets. In 315 women with a singleton gestation receiving 17-OHPC weekly, the concentrations ranged between 3.7–56ng/mL between 25 and 28 weeks of gestation⁵³. Women with 17-OHPC plasma concentrations in the lowest quintile had a higher risk of recurrent spontaneous preterm birth compared to women in the third to fifth quintiles (36.5% vs. 25%, Figure 3). In addition, these women delivered at an earlier gestational age than did women in the third to fifth quintiles. The authors also noted that the lowest rates of preterm birth were seen when 17-OHPC concentration was lower than 6.4ng/mL, but drug concentration and preterm birth risk were not linearly associated (Figure 3)⁵³. Somewhat alarmingly, another report demonstrated an association between 17-OHPC plasma concentrations higher than 14ng/mL and an earlier gestational age at delivery in women carrying twins⁵⁴. The study was unable to uncover the underlying mechanism for the negative association between 17-OHPC supplementation and recurrent preterm birth. The complex nature of the preterm birth phenotype, the lack of knowledge of an exact mechanism and the variation in progesterone receptor polymorphisms limit any conclusion regarding the pharmacodynamic properties of 17-OHPC and the existence of a dose response relationship. In order to characterize the relationship of 17-OHPC and preterm birth, future studies should take into consideration the current knowledge of variation in clinical efficacy and safety of 17-OHPC, identify subsets of women who may most benefit from this therapy, and define groups that may not benefit or may be harmed.

Figure 3. Relationship between median 17-OHPC concentration and preterm birth rate within each quintile.

Relationship between median 17-OHPC concentration from each quintile (Q1–Q5) of 17- OHPC plasma concentrations and the rate of spontaneous preterm birth for each of these quintiles. sPTB: spontaneous preterm birth

Placental transport and fetal exposure to 17-OHPC

Placental transport of 17-OHPC has been demonstrated. At the time of delivery, 1–44 days following the last maternal dose, the range of cord blood 17-OHPC concentration was 0.4–12ng/mL²⁴. Not surprisingly, in specimens collected at the time of delivery, the ratio of fetal to maternal concentrations was 0.2 (range 0.06–0.51) and absolute drug concentrations in the fetus were a function of time since the last dose²⁴.

Fetal hepatocytes were found to metabolize 17-OHPC⁵⁵. The active enzyme was identified as CYP3A7, the fetal form of CYP3A; and the metabolites were noted to be similar but not identical to those resulting from adult hepatocytes⁵⁵. Interestingly, the placenta was also noted to metabolize 17-OHPC to its own distinct metabolites, which differed from those of fetal and adult livers⁵⁶.

Although 17-OHPC is detectable in cord plasma, evidence to date is reassuring in terms of fetal/neonatal safety. Animal studies using doses much higher than those used in humans failed to exhibit any androgenic effect or a risk of congenital anomalies^57–60. A number of authors have also reported the results of well-controlled clinical studies to examine the safety of 17-OHPC in human pregnancy. Varma and Morsman examined the outcome of 150 pregnancies treated with 17-OHPC because of threatened abortion, and compared them with 150 patients who experienced early pregnancy bleeding but were not treated with the drug⁶¹. No evidence was found that the drug had any adverse effect on the fetus or the outcome of the pregnancy. Michaelis, et al. in a cohort study of 13,643 pregnancies in West Germany found no increase in malformations in infants exposed in utero to 17-OHPC compared with controls⁶². Resseguie, et al. examined a cohort of 24,000 pregnancies delivered in Olmstead County, Minnesota from 1936 to 1974 and found that the 649 offspring exposed to 17-OHPC showed no increase in congenital anomalies or other ill effects compared with controls⁶³. A notable feature of this study was the long period of follow-up of the children, with a mean of 11.5 years. Check, et al. performed a follow-up questionnaire study of 382 women treated with progestins during pregnancy and found no increase in anomalies compared with control offspring⁶⁴. Most recently, Northen et. al. evaluated the offspring of women treated with 17-OHPC in the original Meis study⁶⁵. They found no relationship between intrauterine exposure to 17-OHPC and physical findings or developmental characteristics, including gross and fine motor skills, problem solving, and personal/social interactions or gender-specific roles at a mean age of 48 months⁶⁵.

Conclusion

Preterm birth is a complex phenotype with significant morbidity but few successful interventions. Weekly 17-OHPC injections have been proven to decrease rates of recurrent preterm birth. This treatment is associated with limited fetal and maternal risks. However, its efficacy can vary with maternal BMI and in relation to hepatic metabolism and progesterone receptor polymorphisms. While an exact mechanism of progesterone supplementation in preventing preterm birth remains unknown, recent data are shedding light on the variation in outcomes noted in prior clinical studies. The large variation in plasma 17-OHPC concentrations following a fixed dose, and the demonstration of improved efficacy with higher plasma concentrations suggests an opportunity to improve efficacy perhaps by tailoring the dose or by evaluating alternate routes of administration of this essential agent.