Premature Rupture of the Membranes – Treatment consideration

Danforth’s Obstetrics and Gynecology
Ronald S. Gibbs
 
Premature rupture of the fetal membranes (PROM) is one of the most common problems in obstetrics, complicating approximately 5% to 10% of term pregnancies and up to 30% of preterm deliveries. Although the etiology of premature rupture of the membranes is often not clinically evident, a degree of consensus has arisen regarding management options. Indeed, the problem is intricate. Considerations in selecting management in a particular patient are gestational age and patient demographics. The clinician is confronted by a complex set of options including use of corticosteroids, tocolytics, more potent antibiotics, and innovative approaches using various tests (such as amniocentesis, ultrasound, and biophysical testing). Of major importance is the marked improvement in survival of low–birth-weight infants.

PROM is usually defined as rupture at any time before the onset of contractions. Unfortunately, “premature” also carries the connotation of preterm pregnancy. To avoid confusion, we use the word preterm to refer to gestational age of less than 37 weeks. Thus, preterm PROM (PPROM) refers to PROM prior to 37 weeks gestation. The latent period is defined as the time from membrane rupture to onset of contractions. It is to be distinguished from a similar term, latent phase, designating the early phase of labor before the active phase. Various terms have been used to describe presumed maternal or perinatal infections related to PROM. During labor, designations have included “fever in labor,” “intrapartum fever,” “chorioamnionitis,” “amnionitis,” and “intrauterine infection.” The degree of temperature used to define “fever” has varied. After delivery, maternal infection is referred to as “endometritis” or “postpartum infection.” These diagnoses are usually based upon fever, uterine tenderness, and exclusion of other sources of fever. In neonates, the most common term used to report infection was neonatal sepsis, but this may mean strictly a positive blood culture or simply clinical signs or symptoms of sepsis.

The overall approach to management of PROM takes into consideration neonatal survival at the gestational age when rupture occurs. Management may be divided into four different phases of pregnancy. During the second trimester, neonatal survival is nil, leading numerous investigators to adopt a policy of expectant management or induction. Early in the third trimester, neonatal survival rises markedly, but there is still considerable morbidity associated with delivery at this gestational age. In the mid-third trimester, neonatal survival is high, but there is still considerable morbidity, whereas in the late third trimester (at or near term) neonatal mortality and morbidity are low. Neonatal outcome is one of the driving features in determining clinical management.

Diagnosis of Infection After PROM

Both invasive and noninvasive tests have been assessed. As shown in Table 11.1, none of these tests is ideal, particularly because of their low positive predictive values.

Use of Steroids

Clinicians remain unable to agree on the risks and benefits of steroids in preterm PROM. The first meta-analysis in 1989 concluded that the use of steroids “increases the incidence of endometritis and may increase neonatal infections.” The 1994 National Institutes of Health (NIH) Consensus Conference concluded that the risk of maternal and infant infection may be increased with corticosteroid use after PROM but that the magnitude of this risk was small. The NIH recommendations are summarized in Table 11.2. The benefits of steroids with PPROM before 28 weeks, however, have not been firmly established.

In earlier studies, the incidence of amnionitis rose with increasing length of the latent period, but other investigators have found no increase in the incidence of amnionitis among preterm pregnancies with increasing latent periods. In a comparison of outcomes, women with digital examination after PROM had a significantly shorter latent period (2.1 ± 4.0 vs. 11.3 ± 13.4 days; P < .001), more maternal infection (44% vs. 33%; P = .09), and more positive amniotic fluid cultures (11/25 [44%] vs. 10/63 [16%]; P < .05). Thus, routine vaginal examination should be avoided until labor develops in patients with preterm PROM
.
Use of Prophylactic Antibiotics

In patients with PROM prior to term, there are two rationales for prophylactic antibiotics. The first is a clear one; namely, for prevention of perinatal GBS infection
.
A second rationale for antibiotic prophylaxis has been based upon the hypothesis that infection is either the triggering cause of preterm PROM or that infection ensuing after preterm PROM triggers the labor. Accordingly, this rationale for prophylactic antibiotics has been to delay delivery after preterm PROM rather than to prevent clinically evident infection. We believe that good evidence has been provided to favor use of broad-spectrum antibiotics in selected cases of preterm PROM. This support was provided in a meta-analysis and in prospective randomized trials. In the meta-analysis, 24 trials were identified and 13 were included, containing 1,594 women. However, only 6 of the trials were placebo-controlled, and the trials were heterogeneous with regard to antibiotics used. In addition, there was no standard use of steroids, tocolytics, or prophylaxis for GBS. Nevertheless, benefits were demonstrated in favor of women receiving antibiotics. These benefits included a significant delay in delivery within 7 days, a reduction in chorioamnionitis, and a reduction in neonatal sepsis. There were also reductions (that did not achieve statistical significance) in postpartum infection, neonatal death, neonatal pneumonia, and neonatal bacteremia.

In the large MFM trial, patients were enrolled if they had preterm PROM for less than 72 hours at 24 to 32 weeks gestation. Patients were excluded if there was chorioamnionitis, labor, or fetal distress. Patients were then randomized to a course of ampicillin plus erythromycin (each for 2 days i.v. followed by up to 7 days orally) vs. placebo. Patients with GBS were given treatment during the latent period and no tocolytics were used. However, at the time the study was designed, it was decided not to use corticosteroids in any patients. The primary end point was a prospectively defined composite of neonatal death, neonatal RDS, grade III or IV intraventricular hemorrhage, grade II or III necrotizing enterocolitis, or neonatal sepsis. Patients randomized to antibiotic therapy had a significantly greater likelihood of remaining undelivered when assessed at 2 days, 7 days, 14 days, and 21 days. In addition, the primary composite outcome was significantly reduced in the total population and in the GBS-negative cohort. Individual adverse outcomes significantly reduced in the antibiotic group included RDS, chorioamnionitis, neonatal sepsis, and neonatal pneumonia

Subsequent to this trial, others have appeared assessing antibiotics in conjunction with antenatal corticosteroid therapy for patients with preterm premature rupture of the membranes. One study assessed 112 women with PROM from 25 to 35 weeks and randomized them to ampicillin sulbactam/amoxicillin clavulanate versus ampicillin/amoxicillin versus placebo. Tocolytics were used in this trial and betamethasone was used weekly up to 32 weeks. Patients receiving the antibiotics had less serious neonatal complications including neonatal death, RDS, and neonatal sepsis (P < .05) and they also had significantly higher mean birth weight (P = .03). Lewis and colleagues reported a randomized clinical trial of corticosteroids in patients with preterm PROM after treating these patients for a minimum of 12 hours with ampicillin/sulbactam. Antibiotics were continued for 7 days and steroids were repeated weekly. No tocolytics were used. The authors defined the primary outcome as the incidence of RDS, whereas secondary outcome measures included latency and neonatal or maternal infections. In this study of 77 patients, no statistically significant difference in latency was noted comparing the steroid versus no steroid group, and both neonatal and maternal infections were similar. However, there was a significant reduction in the incidence of RDS, 18.4% in the steroid group compared with 43.6% in the no steroid group. The authors concluded that treating preterm PROM with a broad-spectrum antibiotic before corticosteroids decreased RDS without apparent adverse effect. In 1998, a meta-analysis of five trials on antibiotic and glucocorticoid treatment reportedly did not show a significant effect on outcomes including maternal infection, neonatal sepsis, RDS, intraventricular hemorrhage, necrotizing enterocolitis, and neonatal morbidity. In contrast, the authors note “antibiotic therapy without concomitant use of glucocorticoids significantly reduced the odds of maternal infection, neonatal sepsis, and intraventricular hemorrhage substantially.” However, this meta-analysis did not include some of the more recent studies noted immediately above.

In a very large (nearly 5,000 patients) international trial (ORACLE I), patients with PPROM were randomized to one of four courses: oral erythromycin, oral amoxicillin-clavulanic acid, both antibiotics, or oral placebo. Each regimen was taken four times a day for 10 days or until delivery. The primary outcome measure used was a composite of neonatal death, chronic lung disease, or major cerebral abnormality on ultrasound. Erythromycin was associated with several benefits to the neonate (fewer cases with the composite outcome, prolongation of pregnancy, and fewer positive blood cultures). Amoxicillin-clavulanic acid—with or without erythromycin—was associated with prolongation of pregnancy, but it was also associated with a significant increase in neonatal necrotizing enterocolitis. The applicability of this study to contemporary U.S. practice is limited, however, because the authors made no provision for GBS prophylaxis. Other features of the study to emphasize are that antibiotics were used orally, enrollment was permitted up to 37 weeks (only 50% of cases were less than 32 weeks), and there was no standard approach for use of steroids or tocolytics. (Steroids were used in 75% of cases and tocolytics in <15%.)

Widespread use of antibiotics in this situation has raised concern about selection pressure toward resistant organisms, but in the MFM trial, there was no significant increase in maternal yeast infection or neonatal Candida sepsis, nor were there any cases of pseudomembranous colitis, maternal sepsis, or maternal death.

Determination of Fetal Lung Maturity

Because RDS is the single greatest threat to infants with PROM, some investigators have determined the status of fetal pulmonary maturity and proceeded with delivery when there was lung maturity. One study used amniocentesis and obtained fluid in about half of the cases. Others have attempted to collect amniotic fluid from the vagina and have had success rates of 80% to 94%. Presence of either PG or an L/S R of more than two in amniotic fluid collected vaginally has been reported to be a good predictor of pulmonary maturity.

In a larger series of patients with PROM before 36 weeks, investigators determined whether PG was present in the vaginal pool and delivered patients when there was presence of PG, spontaneous labor, or evidence of sepsis. PG in amniotic fluid from the vagina reliably predicted fetal lung maturity. However, absence of PG did not necessarily mean that RDS would develop. Of the 131 patients who did not show PG in the vaginal pool in any sample, 82 (62%) were delivered of infants who had no RDS. Thus, even with PROM, delivery of a premature infant simply because its lungs showed biochemical maturity may be questioned in view of other potential hazards of prematurity and the difficulty of the induction. Of note, some genital tract bacteria have been found to yield a false-positive test for PG.