ORIGINAL ARTICLE

Open randomized comparison of prostaglandin E, given by intracervical gel or vagitory for preinduction cervical ripening and induction of labor HENNING KVISTPOULSEN,' LARSKRAG MOLLER,'JESG. WESTERGAARD,' STENGROVETHOMSEN,~ REYNIR T. GEIRSSON3 AND REYNIRARNGRiMSSON3 From t h e Departments of Obstetrics and Gynecology, 'Esbjerg Hospital, Esbjerg, Denmark, 2 0 d e n s e University Hospital, Odense, Denmark, and 3National University Hospital, Reykjavik, Iceland

Acru 0h.sret Gytiecol S c u d 1991; 70: 549-553

Intr;tcervical application o f prostaglandin E, (PGE?) in a viscous gel was compared with conventional wax-based PGE, vagitories (pessaries) for ripening of the cervix prior to induction of labor. A total of 226 healthy pregnant women at term were randomly allocated t o receive intracervical gel with an effective dose of 0.S mg ( n = 116) o r vagitorics containing 2.5 mg PGEz ( n = 110). All women had a modified cervical score of 5 4. The numbers o f cases contributed by each of the three centers were similar. There was no significant difference in parity, gestational length, maternal characteristics, indications for induction o r preinduction cervical scores between the treatment groups. The rate o f spontaneous birth was 71% in the gel group, compared with 69% in the vagitory group. Successful treatment was defined as active labor within 24 h o r a change in cervical dilatation allowing artificial rupture of the membranes with subsequent progressive labor. The success rate was not significantly different in the gel group (82%) compared with the vagitory group (80%). There were no differences in the frequency of fetal distress, outcome of labor, assisted delivery rates or maternal side effects. The cervical scores werc not different at 12 and 24 h after application. lntracervical gel and intravaginal application o f PGE? were similar in their efficacy and safety for ripening of the cervix and inducing labor at term. K c y words: induction of labor; labor; prostaglandins; randomized clinical trial Siihniirtc~/Jutiuury 4. 1991 Acceptid Jurre IX, 1991

Prostaglandin E? (PGE,) has been used for two decades to promote ripening of the cervix before formal induction of labor by amniotomy and oxytocin administration (1). PGEz has been applied extraamniotically , intracervically and in the vaginal vault with a comparable degree of success in terms of immediate induction of labor o r improvement of cervical score. About 5(k75% of women go into labor following intravaginal, intracervical or extraamniotic application of PGE, (2). Different prepara-

tions of the drug seem to vary in case of administration, incidence of side effects and safety for mother and fetus. Intracervical application in a slowrelease gel preparation has been claimed to be associated with n o or mild uterine activity following kipplication, yet producing a marked improvement in cervical score and increased success of subsequent induction of labor (3). In a multicenter study+ women were randomly allocated into two groups receiving either slow-re-

550

H . Kvist Poidsm el a/.

Table I . Maternal characteristics of women using intracervical PGEz gel or PGEz vagitories

Vagitories"

Geld

Age (years) Parity

Height (cm) Weight (kg) Gcstational length (days) Cervical pretreatment score

Mean

SD

Range

Mean

SD

Range

27.9 1.6 167.3 14.9 282.6 2.2

4.9 0.9 5.8 14.1 11.6 1.2

1743 1-6 15IL180 48-1 I4 250-302 l!-4

28.0 1.7 168.1 77.8 282.6 2.5

5.5 1.1 6.1 17.2 11.8 1.3

1742 1-7 151-186 45-124 259-303

0-4

,'Differences between groups not significant (t-test) lease gel or vagitories (pessaries) to test in a clinical setting the efficacy and safety of these two forms of PGEz administration for inducing cervical ripening and to evaluate the effect of treatment on the course of labor and delivery.

Patients and methods A total of 226 women were randomly allocated to receive either 0.5 mg PGE, in 2 ml of viscous gel applied intraceriically (4) (CerviprostB, Organon, Nethcrlands) or 2.5 mg PGEz in a vagitory placed in the posterior vaginal fornix. The gel was supplied in a prepacked syringe containing the prostaglandin dispersed in dry polydextran (starch polymer) to which immediately before use a solution containing sodium chloride, dextran and water is added in a dual-compartment syringe to produce a viscous gel. This was injected under direct vision into the cervical canal as described earlier (3). The vagitories contained the active substance in a Witepsol S 5 5 wax base and were produced at Odense University Hospital Pharmacy using standard pharmaceutical methods. All women were at term (259 to 303 days menstrual age) as determined from ultrasonic measurements at < 20 weeks' gestation. They were healthy, with singleton pregnancies in a cephalic presentation, with no history of previous cesarean section or cervical conization and not in labor. All had an unfavourable cervix according to a modified cervical score of < 4 points ( 5 ) . Pretreatment evaluation consisted of a general physical examination, basal recordings of blood pressure, heart rate and temperature as well as fetal cardiotocography (CTG) for a minimum of 30 min. Criteria for inclusion or exclusion to the study were predefined, but indications for induction as noted were determined by convention and similar at each of the three centers. After obtaining informed consent, random allocation was by opening a sealed envelope where numArlo Ohster Gyrierol Srand 70 ( I Y Y I )

bering sequence was determined from a randomization schedule constructed to include five gel and five vagitory group allocations in each consecutive group of 10 patients. Following administration, CTG-monitoring was carried out for 3&60 min and repeated at 3-hourly intervals or whenever uterine contractions developed. When labor became established, the membranes were ruptured artificially as soon as possible. Adverse effects were noted. A t 12 h after administration of the drug, the woman was reassessed and if possible the membranes ruptured. Oxytocin infusion (10 IE1500 rnl isotonic saline) was commenced 2 h later when indicated. If rupture of the membranes was not feasible (cervical score 5 4 points), the woman was given the same substance again using the same observation schedule. The cervix was assessed again at 24 h if required. Otherwise, routine labor ward management at the obstetricians' discretion was applied. Information on induction-delivery interval, length of labor, amount of oxytocin, neonatal outcome, birthweight, baseline maternal recordings and adverse effects during labor were noted. All data sheets were computerized at one of the centers (Odense). Successful induction was defined as active uterine contractions with labor progressing within 24 h, and successful priming by whether it was possible to artificially rupture the membranes (ARM) with ensuing labor progress. Failed priminghnduction was considered t o have occurred when cesarean section had to be done without labor having commenced/progressed, the cervical score remaining i 4 (unfavorable for ARM). Length of labor, complications and neonatal data were analysed separately for the success groups. Continuous data were analysed by two-tailed ttests and discrete data by a X*-tcst. Approval was obtained from the medical ethics committee for the detailed study protocol at all centers and approximately an equal third of the women were recruited to each.

Comparative randomized study of PGE,

55 1

Table 11. Outcome of priming and mode of delivery distributed according to number and percentage of total in each group Gel

I n labor/delivered by 12 h In laboridelivered by 24 h (repeated dose) Total in labor/delivered (success) Total not in labor (failure) Cesarean section for fetal distress Cesarean section without further trcatmcnt Repeated priming/stimulation after 24 h

Vagitories

n

%

54 41 9s 21 1 7 13

46.6 35.3 81.9 18.1 0.8 6.0 11.2

n

Y”

SY*

53.6 26.4 80.0 20.0 1.8 7.3 10.9

29 * 88* 22 *

2 8 12*

“Difference not significant (X2-test).

Results Two-hundred and twenty-six women were recruited to the study. Of these, 116 received gel and 110 vagitories. The distribution of women receiving each form of treatment was not significantly different (primiparous 76 gel160 vagitories; parous women 40 ge1/50 vagitories). Maternal characteristics are shown in Table I. These and initial maternal labor ward recordings, as well as individual cervical score components, did not differ significantly between the groups. The main indications for induction were prolonged pregnancy ( n = 36 and 34), hypertensive disease in pregnancy (n = 32 and 27) and intra-uterine growth retardation (n = 11 in each group) with other minor indications evenly distributed between groups. There was no significant difference between maternal characteristics or indications for induction at the three centers. Before treatment commenced all CTG recordings were judged normal. Subsequent to priming, one woman in the gel group and 2 in the vagitory group developed abnormal fetal heart rates and were delivered by cesarean section. In the gel group, all but 16% of the women experienced contractions after

Table 111. Outcome by length of labor (success groups); percentage values are cumulative “el ( n = 95)

Delivery at 5 12 h Delivery at 12-24 h Delivery at 24-36 h Delivery at 3 W 8 h Cesarean section

“Vagitories ( n = 88)

n

Yo

n

%

19 48 11 4 13

23 82 95. 100 14

31

41 89 100

Differences not significant (X2-test).

34 8 -

-

12

14

initial priming, compared with 11% in the vagitory group. Eight in each group developed strong contractions. Four women in the vagitory group developed diarrhea and 3 in each group were nauseous. Two women receiving gel were judged to have uterine hypertonicity, but both gave birth normally without fetal distress being noted. Eight women in the gel group and 9 in the vagitory showed meconium-stained amniotic fluid at amniotomy. There was no difference in analgesic requirement between the groups. A t 12 h, 54 women in the gel group and 59 in the vagitory group were delivered or in labor (difference not significant). The mean cervical score was not different, having increased to 6.1 (SD 2.9) in the gel group and 5.9 (SD 3.0) in the vagitory group. In the former group, 41 women and in the latter 37 were given a repeat dose. The numbers not in labor 12 h after the second dose were 23 and 21 respectively (20% of the whole group). The effect on cervical score was less appreciable after the second application at a mean of 4.35. Outcome is shown in Table I1 and cumulative time of delivery for the success group in Table 111. There was no significant difference between the groups, although somewhat fewer women in the gel group appeared to give birth during the first 12 h. Parous and primiparous women did not differ when the groups were compared and there was no significant difference in the requirements for oxytocin, cesarean section and forcepsivacuum extraction rates, Apgar scores at one and five minutes, blood loss > 500 ml or neonatal illness in the groups. In the gel group, 12 of 21 failed inductions were dealt with by cesarean section, compared with 10 of 22 in the vagitory group. N o fetal death occurred. Times from priming to rupture of membranes (artificial or spontaneous), from priming to established labor and priming to delivery did not differ significantly, even after correcting or parity. When only those delivered vaginally (82 with gel and 76 with Actu OhArer Gynecol Scund 71) (1991)

552

H . Kvist Poirlsen

el al.

vagitories) were considered, there was no difference in any values relating to the length of labor.

Discussion Prostaglandins are a central factor in the control of cervical ripening in the human by acting as chemical messengers having a profound effect on collagen fibril dissociation and ground substance changes leading to increased tissue compliance (6). This effect is also pharmacological (1) and can act separately o n the cervix (7). To what extent it can be dissociatcd from prostaglandin-mediated uterine activity is unknown, Intra-uterine growth retardation may necessitate induction when the cervix is still unfavorable for A R M and does not allow access for internal fetal monitoring. Induction under those circumstances carries a high risk of fetal intra-uterine distress and cesarean delivery. A drug applied locally with a largely isolated effect on the cervix would be desirable in the majority o f growth retardation cases, which tolerate labor well (8). This drug profile has been attributed t o intracervical application of the same PGE, gel preparation as used in this study (3). I n a previous study, intracervical polymer application of 0.5 mg PGE, was found to be more effective than 4 mg of PGEz administered in the same gel intravaginally (3). Only in the group receiving intravaginal gel were gastrointestinal side effects noted, while myometrial activity was significantly less after intracervical gel. Using the same gel and application techniques in a considerably larger group of women. we were unable to confirm those results. Side effects occurred equally in both groups. The only cases of hypertonus occurred in the gel group, but neither of these was of any clinical consequence. In this study a different control substance with a smaller dose of PGEz was used, since this is a more conventional and widely used way of administering PGE,. This may have contributed to fewer side effects in the women using the vagitories. Vaginal administration is, however, associated with systemic prostaglandin absorption soon after administration of even a low prostaglandin dose (9). When the gel is given, the tip of the catheter is advanced gently through the cervical canal into the extra-amniotic space and then withdrawn a short distance before the gel is injected as the catheter is drawn out through the cervical canal. In the confines of the cervical canal it is likely that some of the gel may escape into the extra-amniotic space and come into direct contact with the membranes and decidua, thus leading to increased endogenous prostaglandin E2 and F,, production with enhancement of exogenAcru Obsrvr Gynrcol Sumd

70 (1991)

ous PGEz absorption. This may in practice be a largely unavoidable consequence of the technique and could explain why there was no significant difference in the occurrence of uterine activity following the first dose. Cervical manipulation during the procedures may have contributed to the change in cervical score. The extent of this could not be assessed, but previous studies suggest that this effect is small (1,2,3) with both modes of application used here. Both intracervical and pessary application of PGE2 were shown to be effective and safe. A marked ripening effect occurred in two-thirds of the patients with both drugs after a single application. A further application produced a smaller change in cervical score in both groups, but it has been suggested that the intracervical route may confer some advantage for the group with very low cervical scores (3). The benefit from more than one application may be marginal (10). lntravaginal application requires a higher PGEz dose but is easier than the intracervical route. As the net effect does not differ, the determining factor will be the difference in cost, at present higher for the intracervical preparation.

Acknowledgement We arc grateful for statistical advice from Dr A . Skytthe of the Department of Social Medicine at Odense University.

References I . Calder A A . Cervical ripening. In: Bygdeman M , Berger GS, Keith LG. eds. Prostaglandins and their inhibitors in obstetrics and gynaecology. Lancaster: MTP Press, 1986: 145-64. 2. Shepherd J , Pearce JMF, Sims C D . Prostaglandin vaginal suppositories: a simple and safe approach to the induction of labor. Obstet Gynecol 1981; 58: 596-600. 3. Ekman G , Forman A , Marsiil K, Ulmsten U . Intravaginal versus intracervical application of prostaglandin Ez in viscous gel for cervical priming and induction of labor at term in patients with an unfavourable cervical state. Am J Obstet Gynecol 1983; 147: 6 5 7 4 1 . 4. Wingerup L, Anderson KE, Ulmsten U . Ripening of the cervix and induction of labor in patients at term by single intracervical application of prostaglandin Ez in viscous gel. Acta Obstet Gynecol Scand 1979; Suppl 84: 11-4. 5. Lange AP, Secher NJ, Westergaard J G , Skovgird I . Prelabor evaluation of inducibility. Ohstet Gynecol 1982; 60: 1 3 7 4 7 . 6. Ulmsten U . The cervix. In: Bygdernan M, Berger GS, Keith LG, eds. Prostaglandins and their inhibitors in obstetrics and gynaecology. Lancaster: MTP Press, 1986: 29-57.

Comparative rundomized study of PGE, 7. Ledger WL, Webster M, Harrison LP, Anderson ABM, Turnbull AC. Increase in cervical extensibility during labor after isolation of t h e cervix from the uterus of the pregnant sheep. Am J Obstet Gynecol 1985; 151: 397402. 8. Laurin J , Persson P-H. Polberger S. Perinatal outcome in pregnancies dated by ultrasound. Acta Obstet Gynecol Scand 1987; 66: 337-43. 9. Greer IA, McLaren M. Calder AA. Vaginal administration of PGEz f o r induction of labour stimulates endogenous PGF?,, production. Acta Obstet Gynecol Scand 1990; 69: 621-5.

553

10. Prim RP, Neilson DR. Bolton RN, Mark C, Watson P. Preinduction cervical ripening with sequential use of prostaglandin Ez gel. Am J Obstet Gynecol 1986; 154: 1275-9. Address for correspondence:

Henning Kvist Poulsen, M.D. Department of Obstetrics and Gynecology Esbjerg Hospital DK-6700 Esbjerg Denmark

Acta Obstet Gyiiecol Scand 70 (1991)