Labor Inhibition

With a better understanding of the physiologic regulation of uterine smooth muscle contractility and the availability of many new classes and categories of drugs that affect uterine smooth muscle, it would seem within the obstetrician's grasp to control whether or not the uterus contracts. Uterine smooth muscle stimulants, such as oxytocin and the prostaglandins, have proved to be valuable agents for labor induction and have been shown to be both safe and effective. Labor-inhibiting drugs or tocolytics, however, have been available since 1980, and the incidence of low-birth-weight neonates has increased in the United States.^{1, 2} Of the currently available agents, none has proved to be without significant maternal or fetal side effects. The safety and efficacy of these agents is still questioned, and the exact role of these agents in obstetrics is debated.

Instances in which inhibition of uterine contractions would be appropriate and desirable for good pregnancy outcome include the following: preterm labor, the most widely recognized indication; patient transfer to a tertiary care center; fetal distress associated with contractions; hypertonic uterine contractions; placenta previa when mild bleeding is associated with contractions; and when any surgical procedure is contemplated during pregnancy and carries the risk of preterm labor.

Although the exact sequence of events leading to labor, or preterm labor, is unknown, a great deal is known about the uterine smooth muscle contractile mechanisms. An overview of these mechanisms effectively shows how tocolytic drugs can influence the contractile process.

The smooth muscle cells that make up the myometrium are spindle shaped. They respond to estrogen during gestation by growing and are largest during the later stages of pregnancy.³ Each smooth muscle cell has in its plasma membrane particles that contain the structural and functional proteins involved in the process of ion transport and compose the receptor sites for endogenous and exogenous substances. The cell membrane also contains gap junctions that allow communication from cell to cell and provide synchronization during labor.⁴ Besides containing the normal cellular organelles, the smooth muscle cells of the myometrium contain myofilaments consisting of actin and myosin. The interaction of these two proteins with calcium and adenosine triphosphate (ATP) is the pathway that causes the smooth muscle cell to contract.

The contractile activity of actin and myosin is regulated by the enzyme myosin light-chain kinase, which phosphorylates inactive myosin, causing its activation and interaction with the active site of actin (Fig.1).^{5, 6, 7, 8} The activity of this key enzyme is controlled by the intracellular concentration of ionized free calcium. An increase in intracellular free calcium stimulates calcium binding to calmodulin, which binds to and activates myosin light-chain kinase; conversely, a decrease in intracellular free calcium inhibits calcium binding to calmodulin, inactivating myosin light-chain kinase. Intracellular free calcium is regulated by receptor and voltage-mediated channels in the cell membrane, active Ca²⁺ pumps in the sarcoplasmic reticulum and the cell membrane, the sodium-calcium exchange pump in the cell membrane, and the calcium ATPase system in the cell membrane.⁹

The modulation of myosin light-chain kinase activity by intracellular free calcium has been shown to be linked with hormonal action and can be influenced by pharmacologic agents as well.¹⁰ The myometrial cell membrane contains receptors that convey hormonal or pharmacologic information to the cell interior in one of three ways:

  Type I receptors: Ion channels structured with the receptor as an integral part of the channel, such as the voltage-mediated channels¹¹
  Type II receptors: Receptors structured with an intervening guanosine triphosphate (GTP)-binding protein as a transducer and using inositol triphosphate, diacylglyceride, or arachidonic acid as a second messenger^{12, 13}
  Type III receptors: Receptors structured with enzymatic activity as a part of the receptor mechanism, such as membrane receptor tyrosine kinase, the receptor mechanism for epidermal growth factor¹⁴

Estrogen is necessary for uterine myometrial receptor expression. Both estrogen and progesterone can alter the intracellular calcium store and cause its sequestration or release. Progesterone has been shown to promote Ca²⁺ uptake by the sarcoplasmic reticulum, decreasing intracellular free calcium, and causing myometrial relaxation. ^{15, 16} A metabolite of progesterone has been found to be a τ-aminobutyric acid receptor agonist, potentiating myometrial relaxation.¹⁷

The polypeptide oxytocin is the most specific uterotonic hormone with its own receptor. Expression of the oxytocin receptor in myometrial cells has been detected as early as 12 to 13 weeks' gestation, with receptor concentration increasing up to its maximum concentration at term.^{18, 19} Oxytocin inhibits Ca²⁺ uptake by the sarcoplasmic reticulum¹⁵ and opens the voltage-mediated calcium channels, both processes ultimately aimed at increasing intracellular free calcium and promoting smooth muscle contraction.

Endothelin, a paracrine hormone produced by endothelial cells, acts on adjacent smooth muscle cells and causes vasoconstriction. Endothelin has also been found to cause myometrial contraction equal to oxytocin effect.^{20, 21} Like oxytocin, the myometrial activity of endothelin is dependent on the voltage-mediated calcium channels. Both endothelin and oxytocin have also been shown to stimulate prostaglandin release.^{22, 23}

Prostaglandin (PG) D₂ and I₂ promote uterine smooth muscle relaxation by stimulating their own myometrial receptors. These prostanoid receptors use GTP-binding proteins to activate adenyl cyclase and increase intracellular cyclic adenosine monophosphate (cAMP), thereby reducing intracellular calcium levels.²⁴ PGE₁, PGE₂, and PGF_2α each have been found to have a specific prostanoid receptor and alter the Ca²⁺ permeability of the cell membrane, causing increased intracellular free calcium and smooth muscle contraction.^{25, 26} These prostaglandins also appear to modulate gap junction formation: PGI₂ inhibits gap junction formation, and PGE₁, PGE₂, and PGF_2α promote it.²⁷ Therefore the prostaglandins may represent the final pathway to synchronous uterine contractions. Inhibition of prostaglandin synthesis would restore normal cell permeability, decrease intracellular free calcium, impede gap junction formation, and cause smooth muscle relaxation.

Applying what is known regarding smooth muscle physiology, inhibition of smooth muscle contraction can be effected by agents that (1) increase adenyl cyclase activity, thereby increasing cAMP (e.g., β-adrenergic agonists); (2) compete with oxytocin for the oxytocin receptor, thereby blocking the hormone's action (e.g., oxytocin receptor antagonist); (3) change the intracellular free calcium concentration by inhibiting Ca²⁺ influx at the calcium channels (e.g., calcium-channel blockers, magnesium); or (4) change the permeability of the cell membrane to calcium (e.g., prostaglandin synthetase inhibitors).

The assessment of the safety and efficacy of any new drug is based on carefully conducted and closely controlled clinical trials in humans after extensive drug testing in vitro and in several animal species. Because the leading indication for tocolytic use is spontaneous preterm labor, tocolytics are studied for their safety and efficacy in this population.

Ideally each tocolytic agent should be assessed in a clinical trial randomly assigning patients diagnosed as being in preterm labor to either the therapy under question, an established therapy, or a placebo. In this way a control group is identified that can be used for comparison to the therapeutic group. Blinding the investigators as to which therapy is being used in any clinical setting helps provide unbiased assessment of the agent.

The tocolytic clinical trial begins with a group of patients diagnosed as being in preterm labor. The diagnosis of preterm labor, however, is not easily made. As gestation progresses the uterine cervix normally begins to efface and dilate, and uterine contractions occur more frequently.^{28, 29} Upwards of 40% of women who present with preterm uterine contractions prove to be in false labor and respond to hydration, bed rest, and placebo.^{30, 31} In early clinical trials of tocolytic agents, premature labor was diagnosed by the frequency and duration of uterine contractions without any assessment of the effect of uterine activity on the cervix.³² To define the patient in true premature labor in a more effective way, more recent clinical trials have used criteria to assess the effect of uterine contractions on the cervix. Although specific criteria vary from study to study, in general, preterm labor patients eligible for tocolytic studies should have the following:

  Uterine contraction frequency of more than two contractions in 15 minutes and cervical change under observation
  or
  A cervix dilated 2 to 3 cm on initial examination.

Although tocolytic agents may be less effective with advanced cervical dilatation, there is no evidence that waiting for clinical signs of cervical change compromises tocolytic efficacy.³³

The preterm labor population used to assess tocolytics is not homogeneous, and the etiology of preterm labor leading to preterm birth varies among different populations. Comparing the etiology of low birth weight in two populations of women varying only in insurance status, Meis and colleagues³⁴ found that women receiving public medical assistance had 34% of births resulting from spontaneous preterm labor, 39% of births resulting from premature rupture of the membranes, and 27% of births resulting from medical complications. Women who had private insurance had 59% of births resulting from spontaneous preterm labor, 18% of births resulting from premature rupture of the membranes, and 23% of births resulting from medical complications.

Thus, depending on the population, only approximately 34% to 55% of preterm births would have spontaneous premature uterine activity in which tocolytic agents could be used and might be potentially beneficial. The actual outcome of each pregnancy, however, may be dictated more by the etiology of the premature uterine activity or associated pregnancy complications, or both, than by the use or nonuse of tocolytic agents. For example, in studies of women presenting in spontaneous premature labor with intact membranes, 9% to 20% of patients have been shown to have positive amniotic fluid cultures.^{35, 36, 37, 38} Positive cultures are linked to premature labor at lower gestational ages, more advanced cervical dilatation on admission, and a higher incidence of respiratory distress syndrome and infectious complications in the preterm neonates.³⁵

To assess adequately the impact of a tocolytic agent on a given population, its safety and efficacy must be studied in a population large enough to control for many underlying factors. Studies of spontaneous preterm labor have identified factors associated with preterm labor, but associations do not imply causation.^{39, 40} These factors can be broken down into epidemiologic factors, pregnancy-related factors, underlying maternal factors, and unknown factors (Table 1). In actuality, preterm labor can be thought of as a final common pathway of an often unidentified or uncontrollable process.

TABLE 1. Factors Associated With Spontaneous Preterm Labor

  Epidemiologic factors

  Maternal age
  Race
  Socioeconomic status
  Marital status
  Antenatal care lacking or inadequate
  Short interval between pregnancies

  Pregnancy-related factors

  Multiple gestation
  Polyhydramnios
  Placenta previa
  Placental abruption
  Infection
  Fetal growth retardation
  Genetically abnormal fetus
  Intrauterine fetal demise

  Maternal factors

  Previous preterm birth
  Maternal weight or weight gain
  Smoking, alcohol, drug use
  Psychologic stress
  Coitus
  Chronic disease
  Uterine anomalies
  Cervical incompetence
  Trauma

  Unknown

The safety of a tocolytic agent is based on careful observation of the occurrence and extent of side effects expected from its use as well as the occurrence of unexpected and perhaps untoward events during therapy. Tocolytic safety must assess the effect of the agent on the mother, the fetus, the neonate, and the overall pregnancy outcome. Tocolytic efficacy has been assessed in many clinical trials of tocolytic agents, evaluating whether the therapy prolongs the pregnancy, causes an increase in birth weight in the infants delivered after therapy, or changes the incidence of respiratory distress in infants delivered after therapy.^{41, 42, 43} These assessments of efficacy are relatively easy to measure; however, it has been argued that the best assessment of tocolytic efficacy is measured by a change in perinatal morbidity and mortality.⁴⁴ This assessment of efficacy is not easy to measure and requires large-scale multicentered studies undertaken prospectively to control for extraneous factors that may affect pregnancy outcome (e.g., advances in neonatal intensive care).

To date, tocolytic therapy has fallen short of these goals,⁴⁵ but many agents have been shown to prolong pregnancy long enough to allow maternal administration of antenatal steroids, thereby helping to reduce the incidence of respiratory distress in delivered infants.⁴⁶ Numerous clinical trials of different tocolytic agents have appeared in the literature. Because many studies are from a single center and report on small numbers of patients, care must be taken in evaluating the strength of the evidence presented and in generalizing from small-scale trials to the general population of preterm labor patients.

Skirting many of the issues regarding tocolytic safety and efficacy, health economists have raised a surrogate issue: Is preterm labor therapy cost-effective? Two studies have found that it is less expensive to treat patients in preterm labor before 34 weeks' gestation and prevent preterm birth than to allow preterm labor to progress to delivery.^{47, 48}

Initial Clinical Assessment

Before presenting in premature labor, many women experience one or more of the following: spotting, an increase in vaginal discharge, and crampy lower abdominal pain or pressure.⁴⁹ The presence of these symptoms should be sought during antenatal visits, and any woman complaining of these symptoms should be evaluated for uterine contractions and cervical change. Once a patient presents with painful regular uterine contractions, prompt assessment is necessary because the success of all tocolytic therapy depends on therapeutic intervention before uterine contractions have caused advanced cervical dilatation.^{50, 51} Management of the patient should be dictated by the initial cervical examination. The patient whose cervix is found to be long and closed should be observed over time to assess whether or not contractions continue and produce cervical change. However, the patient presenting with a cervix 3 cm dilated and 50% effaced with regular uterine contractions has a low therapeutic success rate and is best managed by administration of tocolytic therapy as soon as possible. The patient should be placed at bed rest in the left lateral decubitus position; she should undergo fetal and tocodynometric monitoring, intravenous hydration, and serial cervical examinations, preferably by the same examiner. This observation period allows the clinician to distinguish between the 40% to 50% of preterm labor patients who can be managed with bed rest and intravenous hydration alone and who do not progress to delivery or need pharmacologic therapy from the patients who are in true premature labor. Moreover, for patients less than 3 cm dilated, it has not been shown that waiting for cervical change alters the outcome of tocolytic therapy.³³

During this observation period, the patient also should undergo a complete medical and obstetric evaluation because prolongation of pregnancy with tocolytic therapy may be contraindicated. In some patients, pregnancy should not be continued for various obstetric and medical reasons (Table 2), and some clinical situations involve balancing the relative risk of continuing the pregnancy with tocolytic therapy against the risk of preterm delivery. Assessment should include a complete medical and obstetric history and a full physical examination. Ideally, accurate assessment of gestational age of the fetus will be accomplished in early pregnancy, but because adequate prenatal care is neither available to nor sought by all women, many preterm labor patients need ultrasound fetal age assessment. Mindful of the error in fetal age assessment in late pregnancy, many clinicians advocate fetal pulmonary maturity assessment by amniocentesis either before initiating tocolytic therapy or shortly after therapy begins.

TABLE 2. Contraindications to Preterm Labor Inhibition

An underlying infectious etiology for preterm labor must be sought so that diagnosis may be established and appropriate antibiotic therapy initiated. The association of preterm labor with urinary tract infections necessitates a urine examination by both microscope and culture. A meta-analysis has revealed that antibiotic treatment significantly reduces the risk of delivery of a low-birth-weight infant in this population.⁵² Pyrexia associated with infection should be treated because an elevated maternal temperature may itself lead to uterine contractions.⁵³ Cervical cultures for gonorrhea, group B streptococcus, and chlamydia should be obtained and are useful in patient management.

Of particular concern are signs and symptoms of chorioamnionitis, a cause of preterm labor that may be present with intact membranes with no other signs or symptoms except preterm labor.⁵⁴ Routinely available tests for subclinical chorioamnionitis include maternal serum C-reactive protein, a Gram's stain of unspun amniotic fluid, a white blood cell count, and an amniotic fluid glucose concentration. Although C-reactive protein has a low specificity and sensitivity for intra-amniotic infection, its high negative predictive value allows it to exclude amniotic infection as a cause for preterm labor.⁵⁵ A Gram's stain has a low sensitivity, but it is highly specific for intra-amniotic infection and has a high positive and negative predictive value.³⁵ A study of white blood cell counts of amniotic fluid obtained from preterm labor patients has shown that cell counts greater than 50 cells/mm³ have a high sensitivity for intra-amniotic infection, but a low specificity.³⁷ In one study, amniotic fluid glucose concentrations were lower among patients with intra-amniotic infections; amniotic fluid glucose concentrations less than 14 mg/dL had a high sensitivity and specificity for intra-amniotic infections, but had a low positive predictive value.³⁸ Because no one test has been demonstrated to have high specificity and sensitivity with high positive and negative predictive values, combinations of these clinical tests seem appropriate.

Patients whose uterine contractions do not respond to hydration and decreased activity, and whose contractions cause observed cervical change, and who have no contraindications to pregnancy prolongation or to the use of tocolytic agents are candidates for tocolytic therapy.

Pharmacologic Interventions

After the decision to inhibit preterm labor is made, the physician must decide which of the available therapeutic modalities to use. Historically, the obstetrician has used narcotics (e.g., morphine, meperidine) and sedatives as tocolytic agents; however, these agents do not cause smooth muscle relaxation^{56, 57, 58, 59} and really have no place in the therapy of true preterm labor. Intravenous ethanol was used as a tocolytic agent during the 1960s and 1970s, but it has been abandoned because of maternal inebriation and a higher incidence of adverse neonatal outcome.^{60, 61}

Current pharmacologic therapies for preterm labor include the β-sympathomimetics, magnesium sulfate, prostaglandin synthetase inhibitors, slow calcium-channel antagonists, and as a research agent, the oxytocin receptor antagonist atosiban. No agent has been found to be significantly better than all others, and in any individual clinical situation one or another of these agents may be preferable.

b-SYMPATHOMIMETICS.

Pharmacology.

In 1925, Rucker⁶² discovered that small doses of epinephrine inhibited uterine contractions. It was not until 1948, however, that Ahlquist⁶³ demonstrated that this effect was caused by stimulation of β-receptors in uterine smooth muscle. Lands and colleagues⁶⁴ further showed that the β-adrenergic receptors could be divided into two distinct groups: the β₁-receptors, found in cardiac muscle, the small intestine, and adipose tissue; and the β₂-receptors, found in the smooth muscle of the uterus, blood vessels, bronchioles, and diaphragm. Stimulation of β₁-receptors has positive inotropic and chronotropic effects on heart muscle and causes relaxation of intestinal smooth muscle, lipolysis, and glycogenolysis. Stimulation of β₂-receptors relaxes the smooth muscle of the uterus, bronchioles, and blood vessels. Although drugs are classified as having predominantly β₁ or β₂ effects, the difference between β₁- and β₂-receptors is not as distinct as that found between α- and β-receptors, and there is considerable cross-reactivity between drugs classified predominantly as a β₁-agonist or a β₂-agonist.⁶⁵ A drug classified as a β₂-agonist may lose its β₂-specificity at higher doses.

The β-sympathomimetics are all derived from structural modifications of the epinephrine molecule. Increasing the size of the alkyl substituent at the amino group increases β-agonist activity, and substituting a hydroxyl group at the 3 and 5 positions on the aromatic ring confers greater β₂-receptor specificity.⁶⁵ Drugs whose design has been based on these structure-activity relationships and that have been shown to have greater β₂-receptor agonist activity are fenoterol, terbutaline, ritodrine, salbutamol, and hexoprenaline. Clinical studies have demonstrated that all these drugs have tocolytic efficacy,^{66, 67, 68, 69, 70} but in the United States, ritodrine, the only drug approved by the FDA for preterm labor therapy, and terbutaline are the β₂-sympathomimetics commonly used.

The β₂-sympathomimetics cause smooth muscle relaxation by interaction with the β-receptor on the cell membrane. This agonist-receptor complex, through stimulation of a GTP-binding protein activates adenyl cyclase, which converts ATP to cAMP.⁷¹ Increasing intracellular cAMP activates cAMP-dependent protein kinases that cause phosphorylation of a protein associated with the sodium-potassium ATPase system. This in turn increases the rate at which Na⁺ is pumped out of the cell in exchange for K⁺. An increase in the Na⁺ gradient across the cell membrane accelerates the Na⁺-Ca²⁺ exchange, causing a net loss of Ca²⁺ from the cell. The decrease in intracellular calcium deactivates myosin light-chain kinase and causes smooth muscle relaxation. An increase in intracellular cAMP also inhibits myosin light-chain kinase activity directly by causing phosphorylation of the enzyme itself and further decreasing intracellular calcium by promoting Ca²⁺ binding by the sarcoplasmic reticulum. By altering adenyl cyclase activity and the calcium pump, β-adrenergic agonists control the activity of myosin light-chain kinase, the key enzyme controlling the interaction of actin and myosin, thereby influencing the basal contractile state of uterine smooth muscle.^{72, 73, 74}

Maternal, Fetal, and Neonatal Effects.

β₂-sympathomimetics produce a dose-related increase in maternal heart rate and a widening of the pulse pressure.⁷⁵ Tachycardia is caused by both direct stimulation of cardiac β₁-receptors and a reflex baroreceptor activation due to peripheral vasodilation. The lowered diastolic pressure may facilitate venous return to the heart, thereby increasing stroke volume and systolic pressure. Cardiac output has been shown to increase by as much as 56% when β-adrenergic agonists are used.

Maternal metabolic effects caused by the use of β₂-sympathomimetics can be traced to the increased intracellular cAMP and include hyperglycemia from hepatic glycogenolysis, hyperlacticacidemia, lipolysis, and hyperinsulinemia caused by a direct effect of the β-agonist on the pancreatic β-cells.^{76, 77, 78, 79, 80, 81} Because β-sympathomimetics act on the sodium-potassium pump, they also reduce the plasma K⁺ concentration by shifting K⁺ from the extracellular space to the intracellular space.^{82, 83} The hyperglycemia associated with β-sympathomimetics used for tocolysis is seen mainly during intravenous administration of the drugs. The highest glucose concentrations are seen approximately 3 hours after intravenous therapy is initiated, begin to decline thereafter, and return to normal after therapy is stopped.⁸³ Continued oral administration of ritodrine does not appear to affect carbohydrate metabolism adversely,^{84, 85} but oral terbutaline has been shown to impair maternal glucose tolerance and to cause the development of gestational diabetes.⁸⁶ The hypokalemia associated with β-sympathomimetics is seen only during intravenous therapy, and K⁺ concentrations begin to return to normal during the first 24 hours of therapy.⁸³

The β-sympathomimetics also affect total body water by stimulating the renin-aldosterone system and the release of antidiuretic hormone, resulting in a net increase of total body sodium and water.^{87, 88, 89} A study of pregnant baboons revealed that this increase in total body water results in an increase in the interstitial fluid space and does not increase plasma volume.⁹⁰

The use of β-sympathomimetics is associated with the common side effects of nervousness, restlessness, anxiety, tremors, palpitations, nausea, and vomiting, which are caused by sympathomimetic stimulation. The hyperglycemia associated with β-sympathomimetics has precipitated diabetic ketoacidosis when used in diabetics in whom blood sugars were not monitored.^{91, 92, 93} However, the most serious adverse reactions seen with the β-sympathomimetics used for tocolysis relate to effects on the cardiovascular system. A rare maternal death was reported among patients with unrecognized cardiac or pulmonary disease.⁹⁴ Cerebral ischemic episodes have been reported in patients with a history of migraine headaches.⁹⁵ Chest pain or tightness was reported among 1% to 2% of patients treated with intravenous β-sympathomimetics.⁹⁶ Electrocardiographic recordings also were found to exhibit ischemic ST-segment depression on occasion,⁹⁷ but measurements of cardiac enzymes during β-sympathomimetic therapy have not revealed any evidence of myocardial damage.^{98, 99} Cardiac arrhythmias can occur in approximately 2% of patients during intravenous therapy, are usually asymptomatic supraventricular arrhythmias, and respond to discontinuation of therapy.^{100, 101, 102}

Dyspnea occurs in 5% to 12% of patients treated with intravenous β-sympathomimetic agents and may reflect pulmonary congestion.¹⁰³ During the first 24 to 48 hours of therapy, the prevalence of frank pulmonary edema has been reported to occur in 0.3% to 5% of patients.^{104, 105} Patients with multiple gestations who have even higher blood volumes than those with singleton gestations seem particularly vulnerable to the occurrence of pulmonary edema. Originally, glucocorticoids were thought to increase the risk for pulmonary edema, but this observation has not been corroborated in larger studies.¹⁰² The pathophysiology of the pulmonary edema seen with β-sympathomimetic use is poorly understood. Left ventricular failure associated with iatrogenic fluid overload, prolonged exposure to a high dose of a sympathomimetic drug, and increased pulmonary capillary permeability coupled with lowered intravascular colloid osmotic pressure are tentative explanations. Hemodynamic monitoring of a few such patients with Swan-Ganz catheters has revealed that this pulmonary edema may not represent congestive heart failure because pulmonary wedge pressures have been found to be normal.¹⁰⁶ This lends support to the hypothesis that the mechanism of action is noncardiac pulmonary edema caused by altered pulmonary capillary permeability, similar to the pulmonary edema caused by epinephrine and norepinephrine. However, a controlled experimental study of ritodrine in baboons did not reveal pulmonary capillary membrane changes.¹⁰⁷ In another study, noninvasive techniques in humans revealed a progressive elevation in pulmonary wedge pressures, suggesting that left ventricular failure is the pathophysiologic cause of the pulmonary edema.¹⁰⁸

Fetal effects of β-sympathomimetics can be traced to placental transfer of the active drug or to altered maternal status. Both terbutaline and ritodrine cross the placenta, with ritodrine passing more readily.¹⁰⁹ During intravenous therapy with β-sympathomimetics, a mild fetal tachycardia and increased fetal heart rate variability can be seen.^{110, 111} Fetal breathing movements also increase with terbutaline therapy.¹¹²

Infants whose mothers were treated with subcutaneous terbutaline within 2 days of delivery can develop transient hypoglycemia associated with elevated cord insulin levels.¹¹³ Fetal cardiac effects reported include myocardial necrosis associated with prolonged antenatal use of terbutaline,¹¹⁴ supraventricular tachyarrhythmias causing fetal hydrops,^{115, 116} and ventricular septal thickening.¹¹⁷ In prospective controlled studies of ritodrine, however, these effects on the neonate have not been seen. A large-scale retrospective analysis controlling for many confounding variables demonstrated an association between maternal use of a β-mimetic agent for preterm labor tocolysis and the occurrence of neonatal periventricular-intraventricular hemorrhage.¹¹⁸ Use of a β-mimetic was associated with more than a twofold increase in risk of neonatal periventricular-intraventricular hemorrhage, with the same increase in risk for the occurrence of grades 3 and 4 periventricular-intraventricular hemorrhage. Nevertheless, a neonatal outcome study and follow-up studies from 1 to 9 years of infants born to women treated with β-sympathomimetics have found no demonstrable abnormalities in condition at birth or overall growth and development.^{119, 120}

Usage Guidelines.

After a patient has been assessed for any contraindications to preterm labor inhibition, use of a β-sympathomimetic requires that the patient be assessed for any absolute or relative contraindications specific to their use (Table 3). Baseline maternal vital signs, a 20- to 30-minute fetal monitor strip, blood chemistries, and an electrocardiogram should be obtained before therapy is begun. Treatment with ritodrine initially is begun by intravenous administration with a controlled infusion device. Terbutaline has been shown to have unacceptable cardiovascular side effects when administered intravenously, and this mode of administration has been abandoned.¹²¹ Subcutaneous administration of terbutaline has been reported, however, with minimal maternal and fetal side effects.¹²²

TABLE 3. Recommended Contraindications to Use of β—Μimetic Drugs

Used for uterine tocolysis, ritodrine should be begun only after initial intravenous hydration because the β-sympathomimetics cause peripheral vasodilation and can cause hypotension. A urine-specific gravity test can be used as a rough assessment of hydration status in a given patient. A urine-specific gravity greater than 1.020 would dictate a greater fluid need, and 1000 mL should be infused over a period of 30 to 60 minutes. A more dilute urine, with a specific gravity less than 1.020, would dictate less fluid need, and 500 mL would be infused over the same time period.

Ritodrine is administered diluted in normal saline at an initial dose of 0.05 mg/minute. The dose is increased at 10- to 30-minute intervals until contractions cease or are significantly reduced or until unacceptable side effects develop. Before the infusion rate is increased, the maternal pulse and blood pressure should be monitored, and uterine activity and fetal heart rate should be noted. Because β-sympathomimetics cause salt and water retention, the mainline intravenous infusion rate should be adjusted with each increase in dose so that total intravenous fluids do not exceed 100 to 120 mL/h or 2500 mL during a 24 hour period. Tocolytic therapy with β-sympathomimetics is titrated not only to suppress uterine activity, but also to maintain maternal pulse rate and blood pressure within acceptable limits. The maternal pulse should increase by approximately 20 to 40 beats/minute over the baseline pulse rate. If no maternal pulse increase is noted, the drug dose can be increased until contractions cease, even if recommended maximum doses are exceeded. The dose should not be increased, however, if maternal pulse rate is 130 beats/minute or more. A widening of the pulse pressure also can be expected, which can cause a significant decrease in blood pressure. These cardiovascular changes can be reversed by a decrease in the ritodrine infusion rate or by an increase in intravenous fluids. Care must be exercised, however, not to overload the patient with fluids.

Once uterine tocolysis is achieved, the infusion is continued for 6 to 24 hours. The manufacturer recommends that the ritodrine infusion be continued at the infusion rate at which uterine relaxation occurred.¹²³ Caritis and colleagues¹²⁴ made an alternative recommendation. Because the manufacturer's regimen can result in excessively high plasma concentrations of ritodrine and may be related to adverse maternal cardiovascular side effects, they suggest that, after uterine tocolysis has been achieved, the ritodrine infusion rate be reduced to the lowest infusion rate that maintains continued uterine relaxation and that the infusion rate be continued for 12 hours.

An alternative regimen to the use of intravenous ritodrine has been described in which a dose of 0.25 mg terbutaline is administered subcutaneously and repeated hourly, if maternal pulse rate is less than 120 beats/minute, until uterine contractions cease.¹²² Terbutaline is then continued orally in a dose of 5 to 10 mg every 4 hours. Efficacy in a retrospective review was found to be comparable to the efficacy of ritodrine. Subcutaneous terbutaline also has been used to decrease myometrial activity and to increase uteroplacental blood flow in the therapy of fetal distress.^{125, 126}

Oral medication is begun 30 minutes before the intravenous infusion is discontinued. Oral therapy should be titrated to sustain an increase in maternal pulse rate, which should remain between 100 and 120 beats/minute, or 20 to 40 beats/minute over the baseline maternal heart rate. A comparison of oral ritodrine and terbutaline indicates that terbutaline may be more effective in preventing recurrent labor.¹²⁷ The need for continued oral therapy, however, has been questioned by many. The original studies that showed a decrease in the incidence of recurrent episodes of preterm labor used uterine contractions, not cervical change, as the criterion for preterm labor.¹²⁸ When stricter criteria for establishing the diagnosis of preterm labor were used in a randomized clinical trial, oral terbutaline was not shown to reduce the rate of preterm birth.¹²⁹

If uterine contractions recur, the maternal pulse rate will indicate whether a supplemental oral dose can be given safely in an attempt to restore tocolysis without resorting to intravenous therapy. With continued contractions and demonstrated cervical change, intravenous therapy can be reinstituted if indicated in the overall management of the preterm labor patient.

Clinical Efficacy.

Evidence of the clinical efficacy of the β-sympathomimetics in the therapy of preterm labor has been provided by prospective clinical trials of ritodrine and terbutaline, which have demonstrated their ability to suppress uterine contractions and delay delivery.^{41, 42, 50, 130} Initial trials, however, used varying protocols of administration at different centers, an inadequate number of controls, and nonstandardized treatments in the control population. Subsequent randomized trials of ritodrine and terbutaline, both of which seem to have equal efficacy in arresting preterm labor, have failed to show a benefit of these β-sympathomimetics over placebo in the long-term prolongation of pregnancy.^{104, 131, 132, 133, 134, 135, 136} Critical assessment, including a meta-analysis of existing clinical trials of the β-sympathomimetics, has shown that β-sympathomimetics can delay delivery for at most 24 to 48 hours and that the drugs have no effect on perinatal morbidity or mortality.¹³⁷ The clinical value of the β-mimetics, however, may be in that they provide a delay in delivery and allow administration of corticosteroids and transfer to a tertiary care center for delivery.

Use of the β-sympathomimetics in patients with premature rupture of the membranes has been studied in three prospective randomized studies.^{138, 139, 140} The overall weight of the clinical evidence is that their use resulted in no significant difference in delay of delivery or neonatal outcome between the treated group and the control group, and their use was therefore of no benefit.

MAGNESIUM SULFATE.

Pharmacology.

Essential as a cofactor for many cellular enzyme systems and in neuromuscular transmission, magnesium is a naturally occurring intracellular cation. In the 1950s Hall and colleagues¹⁴¹ observed that elevated levels of Mg²⁺ reduced uterine contractility in vitro and prolonged labors in patients treated with magnesium sulfate for preeclampsia. Magnesium sulfate alters the extracellular and intracellular Mg²⁺ concentration, which in turn inhibits Ca²⁺ entry through calcium channels into the myometrial cell.^{142, 143} Magnesium sulfate does not appear to effect Ca²⁺ release from intracellular stores. The net effect of magnesium sulfate on uterine smooth muscle is a decrease in intracellular Ca²⁺ and inhibition of myosin light-chain kinase and the binding of actin and myosin.

The normal plasma concentration of Mg²⁺ is 1.5 to 2.2 mEq/L. Mg²⁺ is eliminated by glomerular filtration through the kidney and is reabsorbed in the proximal renal tubules.¹⁴⁴ Serum Mg²⁺ concentrations have been reported to decrease during early pregnancy, to increase gradually throughout the second and third trimesters, and to decrease again immediately before delivery.¹⁴⁵

A high plasma Mg²⁺ concentration is toxic and is rarely seen in persons with normal renal function. Elevated levels of Mg²⁺ depress neuromuscular function by inhibiting acetylcholine release at the motor nerve end plate. Clinically this is manifested by loss of deep tendon reflexes. Although there have been reports of loss of deep tendon reflexes at Mg²⁺ plasma levels as low as 4 mEq/L, generally deep tendon reflexes are not lost until plasma concentrations reach levels of 8 to 10 mEq/L.¹⁴⁶ Respiratory depression occurs with plasma concentrations of 12 to 15 mEq/L.^{146, 147} At plasma levels of 10 to 15 mEq/L, Mg²⁺ prolongs cardiac conduction time as well as the PR and QRS intervals. Plasma concentrations greater than 15 mEq/L cause cardiac arrest.¹⁴⁸

Maternal, Fetal, and Neonatal Effects.

Common maternal side effects of magnesium sulfate therapy include flushing, a warm sensation, headache, lethargy, dizziness, nausea, and blurred vision. Cardiovascular effects include a decrease in peripheral vascular resistance, a small increase in cardiac output, a mild increase in heart rate, and a widening of the pulse pressure.¹⁴⁸ From animal studies, magnesium sulfate has been shown to produce a mild increase in uteroplacental blood flow and to maintain perfusion pressure.^{149, 150} Pulmonary edema has been reported in approximately 1% of patients treated with magnesium sulfate therapy for preterm labor.¹⁵¹ During magnesium sulfate therapy, maternal Ca²⁺ levels decrease by approximately 25%, which is caused by increased urinary Ca²⁺ excretion.^{152, 153} Maternal bone density has been shown to decrease in magnesium sulfate-treated patients compared with controls.¹⁵⁴

Distributed through total body water, Mg²⁺ readily crosses the placenta, with fetal and neonatal levels proportional to maternal levels.¹⁵² Mg²⁺ has been shown not to affect fetal heart rate variability,^{155, 156} but it may cause a nonreactive nonstress test and a decrease in fetal breathing movements.¹⁵⁷ Immediate neonatal effects include respiratory depression and decreased muscular tone; generally one to two Apgar points are deducted for decreased tone and drowsiness.¹⁵⁸ These adverse effects most likely are related to the Mg²⁺ umbilical cord level at delivery and the duration of magnesium sulfate therapy.^{159, 160, 161, 162, 163} Prolonged magnesium sulfate infusions generally begun during the second trimester of pregnancy have been shown to cause demineralization of the fetal long bones, rachitic bony abnormalities, and dental enamel hypoplasia.¹⁶⁴ Bony abnormalities can resolve during the neonatal period, but dental enamel hypoplasia may persist.¹⁶⁵

Usage Guidelines.

Used as a tocolytic agent, the empiric therapeutic range of Mg²⁺ is 4 to 8 mEq/L,^{158, 166} which is achieved after a loading dose of 4 to 6 g of magnesium sulfate infused for a 20 minute period, followed by infusion rates of 1 to 3 g of magnesium sulfate every hour, but higher infusion rates may be necessary. Tocolysis with magnesium sulfate has not always been correlated with serum levels, and barring clinical indication of toxicity, higher infusion rates have been reported.¹⁶⁷ Total parenteral fluids should be controlled strictly during therapy because overhydration is the mechanism by which magnesium sulfate causes pulmonary edema.¹⁶⁸ Monitoring of the patient during therapy includes periodically checking for deep tendon reflexes, monitoring urine output, and measuring plasma Mg²⁺ concentrations, especially in patients with altered renal function or decreased urinary output. Usually magnesium sulfate is infused for 12 to 24 hours, but some patients require longer periods of treatment. Because magnesium sulfate is not tolerated orally, many clinicians use oral β-sympathomimetics for long-term management. Alternatively, magnesium gluconate or oxide, magnesium salts with fewer gastrointestinal side effects, can be given in a dose of 1 g every 2 to 4 hours and are reportedly as effective as ritodrine with less adverse side effects.^{169, 170, 171}

Clinical Efficacy.

The clinical efficacy of magnesium sulfate first was reported in an observational study by Spisso and co-workers,¹⁷² which showed that magnesium sulfate can delay delivery in preterm labor patients with intact membranes for 48 hours in 71% of patients, for more than 7 days in 45% of patients, and until term in 31% of patients. Elliott,¹⁷³ reporting on a retrospective analysis of preterm labor patients treated with magnesium sulfate, found that the greatest success could be achieved in singleton pregnancies with intact membranes and that therapeutic success was directly related to the degree of cervical dilatation on admission. Labor was arrested for 48 hours in 87% of patients with cervical dilatation of 2 cm or less, in 62% with dilatation of 3 to 5 cm, and in 31% with dilatation of 6 cm or more.

The efficacy of magnesium sulfate was first shown in a randomized comparison trial using intravenous ethanol and dextrose as control therapies.⁴³ This study indicated that magnesium sulfate was more effective than alcohol in inhibiting uterine contractions for 24 hours, and therapeutic success was correlated with the degree of cervical dilatation at the time treatment was begun. Cotton and associates¹³⁴ reported on a clinical trial randomizing preterm labor patients to magnesium sulfate therapy, terbutaline therapy, or placebo; this study found no difference among these modalities in their ability to delay delivery for 48 hours. The largest clinical trial included 156 women randomized either to magnesium sulfate or to placebo.¹⁷⁴ Criteria for admission into the study did not include cervical change under observation, and 42% of the treated group and 64% of the control group delivered more than 1 week after treatment. Magnesium sulfate was found to have no significant effect on duration of gestation, birth weight, neonatal morbidity, or perinatal mortality. Mean delay of delivery for the treated group was 26 hours versus 22 hours for the control group. Reports from clinical trials comparing magnesium sulfate to β-sympathomimetics have shown that magnesium sulfate is as effective as the β-sympathomimetics in suppressing preterm labor, but magnesium sulfate has fewer maternal and fetal side effects.^{121, 175, 176, 177, 178} These findings have prompted the suggestion that magnesium sulfate be used as the primary tocolytic agent and that ritodrine be reserved for patients whose contractions cannot be inhibited with magnesium sulfate. Magnesium sulfate is also probably safer for use if fetal status is at all compromised because it does not impair uteroplacental perfusion.¹⁵⁰

In attempts to improve clinical efficacy, magnesium sulfate has been combined with ritodrine in patients whose labors could not be arrested with ritodrine alone. In one randomized clinical trial of combined magnesium sulfate and ritodrine compared with magnesium sulfate alone, combined therapies produced mild improvement in tocolytic efficacy but were associated with a greater number of cardiovascular side effects.¹⁷⁹ Hatjis and colleagues¹⁸⁰ reported on combined therapy of ritodrine plus magnesium sulfate in a highly select group of patients failing ritodrine tocolysis alone and showed a 60% success rate, with no increase in maternal or fetal complications. In a randomized clinical trial comparing the efficacy of ritodrine alone to the combination of ritodrine plus magnesium sulfate, ritodrine plus magnesium sulfate was found to be more efficacious than ritodrine alone and did not appear to increase the frequency of adverse side effects.¹⁸¹

The majority of clinical trials on magnesium sulfate therapy of preterm labor therefore show that magnesium sulfate therapy has similar efficacy to the β-sympathomimetics and at best delays delivery for 24 to 48 hours.

PROSTAGLANDIN SYNTHETASE INHIBITORS.

Pharmacology.

Prostaglandins stimulate the intracellular influx of Ca²⁺ and cause release of Ca²⁺ from the sarcoplasmic reticulum.^{182, 183} Prostaglandins also enhance myometrial gap junction¹⁸⁴ formation and may represent the final common biochemical mediator leading to uterine contractions and labor. Nonsteroidal anti-inflammatory agents, such as aspirin, indomethacin, ibuprofen, and naproxen, have been shown to inhibit the enzyme cyclo-oxygenase, which converts arachidonic acid to prostaglandins in the first step of the prostaglandin cascade, hence the name prostaglandin synthetase inhibitors. Of all the anti-inflammatory agents, aspirin is the only one that inhibits cyclo-oxygenase irreversibly.¹⁸⁵ Of the available agents, indomethacin is the agent that has been most studied as a tocolytic agent.

Indomethacin is a potent inhibitor of cyclo-oxygenase, having anti-inflammatory, analgesic, and antipyretic activity. Renal effects of prostaglandin synthetase inhibitors include a reduction in glomerular filtration rate, reduced sodium and water clearance, and lower plasma renin activity, making their use inadvisable in patients with renal disease.¹⁸⁵ Indomethacin has been reported to cause neutropenia, thrombocytopenia, and in rare instances aplastic anemia. Platelet function is altered by inhibition of platelet agglutination, causing a prolonged bleeding time, and peripartum use has been associated with postpartum hemorrhage.¹⁸⁶ Gastrointestinal side effects include anorexia, nausea, abdominal pain, and ulcerative lesions of the gastric and intestinal mucosa. Rare, but fatal, cases of hepatitis have been reported. The central nervous system can be affected most commonly by severe frontal headaches, but dizziness, depression, and psychosis have been seen with prolonged use of indomethacin.¹⁸⁵

Indomethacin, whether administered orally or rectally, is absorbed rapidly and completely; peak plasma levels occur 1 to 2 hours after administration.¹⁸⁷ Time to peak indomethacin serum concentration after oral administration occurs less rapidly than after rectal administration, and peak serum concentration can be further delayed by ingestion with food.¹⁸⁷ To minimize gastrointestinal discomfort from the drug, indomethacin can be administered with food. Peak plasma concentrations of indomethacin given orally to laboring patients are reached somewhat later and are lower than plasma concentrations found in nonpregnant women.¹⁸⁸ Studies in nonpregnant adults show that the half-life has great biologic variation, ranging from 2.6 to 11.6 hours, with a mean half-life of 5.8 hours, hence the use of a 6-hour dosing interval.¹⁸⁷ Indomethacin, largely eliminated through hepatic metabolism in nonpregnant women, is eliminated unmetabolized during pregnancy.¹⁸⁸ The drug is 90% bound to albumin and binds extensively to tissue.¹⁸⁹

Maternal, Fetal, and Neonatal Effects.

In clinical trials of indomethacin for preterm labor, maternal side effects have been rare. Reported side effects of indomethacin include platelet function abnormalities, producing a prolonged bleeding time and an increased risk of postpartum hemorrhage; acute transient renal insufficiency¹⁹⁰; gastrointestinal nausea, vomiting, and diarrhea; and headaches and dizziness.¹⁹¹ Patients with a history of aspirin allergy should not receive other nonsteroidal anti-inflammatory drugs because they have a high cross-reactivity with aspirin allergy. Other medical conditions that contraindicate their use include gastrointestinal ulcerative disease, hepatic or renal dysfunction, and coagulopathies.

Indomethacin is known to cross the placenta.¹⁹² Of greatest concern, explaining the lack of widespread acceptance of indomethacin as a tocolytic agent, are the reported effects of indomethacin on the fetus and neonate. Prostaglandin synthetase inhibitors, especially aspirin, have been reported to cause premature closure of the ductus arteriosus in the fetus and have been associated with pulmonary hypertension as well as persistent fetal circulation in the neonate.^{193, 194, 195} Doppler ultrasound studies of the fetal ductus arteriosus during indomethacin treatment of the mother have revealed transient ductal constriction.^{196, 197} The frequency of ductal constriction is greater with increasing gestational age, occurring in 60% of fetuses between 31 and 34 weeks' gestation, in 40% of fetuses between 27 and 30 weeks' gestation, and in 0% of fetuses less than 27 weeks' gestation.¹⁹⁸ Ductal constriction also is associated more with long-term use of prostaglandin synthetase inhibitors than with short-term use.^{199, 200}

Recently an increased incidence of other neonatal complications that could be related to antenatal use of indomethacin has been reported. These complications include bronchopulmonary dysplasia,²⁰¹ necrotizing enterocolitis, grades II to IV intracranial hemorrhage, and persistent ductus arteriosus.²⁰² Because none of these complications have been noted in any of the controlled clinical trials, and because these complications also are seen secondary to prematurity, corroboration from other studies is necessary.

Also observed with use of prostaglandin synthetase inhibitors is oligohydramnios caused by a decrease in fetal urinary output occurring within 5 hours of indomethacin treatment.^{203, 204} This effect of indomethacin has been associated with perinatal death and neonatal anuria.^{205, 206}

Usage Guidelines.

Used as a tocolytic, an initial loading dose of 50 to 100 mg indomethacin can be administered either orally or as a rectal suppository, followed by 25 mg given orally every 6 hours. Concern over the effects of indomethacin on the fetus correlated to gestational age and prolonged use has prompted most clinicians to limit use of indomethacin to pregnancies less than 32 weeks' gestation and for periods of 24 to 48 hours, although repeated courses or longer courses of therapy have been reported. Ultrasound assessment of amniotic fluid volume and Doppler assessment of the ductus arteriosus is recommended after 48 to 72 hours of use.

Clinical Efficacy.

The efficacy of indomethacin as a tocolytic agent was first reported in an uncontrolled observational study.²⁰⁷ This study showed that indomethacin arrested uterine contractions in 80% of patients within 2 hours of treatment. As with other tocolytic agents, therapy was more successful if the cervix was dilated less than 3 cm, delivery being postponed until a birth weight of greater than 2500 g was achieved in 89% of these patients. In a large observation study, Dudley and associates²⁰⁸ reported on a series of 167 patients treated with indomethacin for preterm labor. Of these 167 patients, 79% had delivery postponed for more than 72 hours, and 67% had delivery postponed for more than 1 week. In the infants delivered after therapy, there were no cases of premature closure of the ductus, persistent fetal circulation, or neonatal bleeding disorders.

Niebyl and co-workers conducted a prospective randomized placebo-controlled study of oral indomethacin administered for a 24 hour period as therapy for preterm labor.²⁰⁹ Indomethacin was found to be considerably better than placebo and delayed delivery for 24 hours in 80% of patients. No evidence of premature closure of the ductus or pulmonary hypertension was found in the infants of the indomethacin-treated group. In a second randomized clinical trial of indomethacin compared with placebo, a 7-day delay of delivery was seen in 83% of indomethacin-treated patients compared with 17% of placebo-treated patients.²¹⁰

Trials comparing indomethacin with ritodrine have shown equal efficacy, but fewer maternal side effects in the indomethacin-treated groups.^{200, 211, 212} These trials also first reported serious neonatal side effects, including pulmonary hypertension, bronchopulmonary dysplasia, and necrotizing enterocolitis occurring in controlled clinical trials. Indomethacin also has been compared with magnesium sulfate and has exhibited equal clinical efficacy, but fewer maternal side effects.²¹³

Currently the place of indomethacin in the obstetrician's armamentarium of tocolytic drugs remains controversial. However, in carefully selected patients in whom standard tocolytic agents have not succeeded in producing uterine relaxation and fetal maturity has not been achieved, especially if the fetal gestational age is less than 28 weeks or an immature amniotic fluid lecithin/sphingomyelin ratio has been shown, use of short-term indomethacin is more than justified.²¹⁴ Indomethacin is possibly also the drug of choice for patients whose pregnancy is complicated by polyhydramnios and preterm labor.²¹⁵

CALCIUM-CHANNEL BLOCKERS.

Pharmacology.

Primarily affecting the voltage-dependent calcium channels within the cell membrane and inhibiting the influx of Ca²⁺ into the cell, the slow-channel calcium blockers or calcium antagonists inhibit smooth muscle contraction.²¹⁶ Within the cell, the calcium-channel blockers also inhibit release of Ca²⁺ from the sarcoplasmic reticulum and promote Ca²⁺ efflux from the cell.²¹⁷ These slow-channel calcium blockers are used primarily in the treatment of cardiac arrhythmias, ischemic heart disease, and hypertension. Verapamil was first examined as a tocolytic agent and was shown to impair atrioventricular conduction before effecting uterine tocolysis.²¹⁸ Nifedipine, however, has proved effective as a tocolytic agent in small clinical trials.^{219, 220, 221}

Nifedipine is absorbed rapidly after oral administration.²²² Sublingual absorption is incomplete and actually is caused by swallowing and gastrointestinal absorption. In the preterm patient, peak serum concentrations after sublingual administration show wide variation and a mean half-life of 81 minutes.²²³ Placental transfer occurs with a fetalto-maternal concentration ratio of 0.93 ± 0.2.²²⁴

Maternal, Fetal, and Neonatal Effects.

Maternal treatment with nifedipine causes peripheral vasodilation with small decreases in diastolic blood pressure and hemodilution.²²⁴ Symptoms reported with use include nausea, flushing, headache, and dizziness. A rare case of hepatotoxicity has been reported with maternal use for premature labor.²²⁵ Concomitant use of nifedipine and magnesium sulfate has produced potentiation of hypotension and neuromuscular blockade.^{226, 227}

Nifedipine has been shown to cause a decrease in uterine blood flow accompanied by in utero fetal hypoxemia, fetal acidosis, and fetal death in sheep, goats, and monkeys given calcium-channel blockers.^{228, 229, 230, 231} However, in the small clinical trials using nifedipine for preterm labor, no a

If uterine relaxation is achieved in the preterm labor patient, management includes decreased maternal activity, generally involving bed rest. Depending on gestational age, maternal administration of steroids for fetal benefit should be strongly considered. Whether continued hospital care is necessary must be decided on an individual basis: Cervical dilatation, degree of uterine activity, maternal anxiety, maternal understanding and compliance with therapy, and distance from home to hospital all should be taken into consideration.

If initial attempts at uterine tocolysis fail and labor seems progressive, the obstetrician should consider whether maternal transport to a perinatal and neonatal intensive-care facility is possible and in the best interests of the mother and neonate.