Progesteron und Cerclage–fürwen? · ALLGEMEINES KRANKENHAUS Die menschliche Größe DER STADT WIEN undfeto-maternale Medizin ABTEILUNG FÜR Progesteron und Cerclage–fürwen?

ALLGEMEINES

KRANKENHAUS

DER STADT WIENDie menschliche Größe

und feto-maternaleMedizin

Geburtshilfe

Progesteron und Cerclage – für wen?

Christof Worda

Universitätsklinik für Frauenheilkunde WienAbteilung für Geburtshilfe undfeto-maternale Medizin

ALLGEMEINES

KRANKENHAUS



ABTEILUNG FÜRGeburtshilfeProgesteron und Cerclage – für wen?

5 typische Konstellationen

ü Patientin mit 1 Frühgeburt in der Anamneseü Patientin mit mit 2 oder mehr Frühgeburten in der Anamneseü Patientin mit einer Frühgeburt und verkürzter Cervixü Patientin ohne Frühgeburt und verkürzter Cervixü Patientin mit aufgebrauchter Cervix und Fruchtblase sichtbar

ALLGEMEINES

KRANKENHAUS




Bedeutung der Frühgeburt

ü 1 von 8 Schwangeren entbindet vor der SSW37ü Diese Frühgeburten sind für 85% der Morbidität und

Mortalität verantwortlich

Alexander et al. 2003

ALLGEMEINES

KRANKENHAUS




Dodd et al., 2013

Cochrane Database of Systematic Reviews

Prenatal administration of progesterone for preventingpreterm birth in women considered to be at risk of pretermbirth (Review)

Dodd JM, Jones L, Flenady V, Cincotta R, Crowther CA

Dodd JM, Jones L, Flenady V, Cincotta R, Crowther CA.

Prenatal administration of progesterone for preventing preterm birth in women considered to be at risk of preterm birth.

Cochrane Database of Systematic Reviews 2013, Issue 7. Art. No.: CD004947.

DOI: 10.1002/14651858.CD004947.pub3.

www.cochranelibrary.com

Prenatal administration of progesterone for preventing preterm birth in women considered to be at risk of preterm birth (Review)

Copyright © 2015 The Cochrane Collaboration. Published by John Wiley & Sons, Ltd.

Insgesamt 11 Arbeiten mit 1899 Frauen eingeschlossen (hohes Risiko für eine Frühgeburt)

Progesteron reduziert bei St.p. Frühgeburt oder verkürzter Cervix:

● Geburt <34 SSW (relative risk [RR] 0.31,95%CI 0.14-0.69) ● Geburt <37 SSW (RR 0.55,95%CI 0.42-0.74)

● Neonataler Tod (RR 0.45,95%CI 0.27-0.76)

● Künstliche Beatmung (RR 0.40,95%CI 0.18-0.90)

● Nekrotisierende Enterokolitis (RR 0.30,95%CI 0.10-0.89)

● NICU Aufnahme (RR 0.24,95%CI 0.14-0.40)

ALLGEMEINES

KRANKENHAUS




Meis et al., 2003

n engl j med

348;24

www.nejm.org june

12, 2003

2379

The

new englandjournal

of

medicine

established in 1812

june

12

,

2003

vol. 348 no. 24

Prevention of Recurrent Preterm Deliveryby 17 Alpha-Hydroxyprogesterone Caproate

Paul J. Meis, M.D., Mark Klebanoff, M.D., Elizabeth Thom, Ph.D., Mitchell P. Dombrowski, M.D., Baha Sibai, M.D., Atef H. Moawad, M.D., Catherine Y. Spong, M.D., John C. Hauth, M.D., Menachem Miodovnik, M.D.,

Michael W. Varner, M.D., Kenneth J. Leveno, M.D., Steve N. Caritis, M.D., Jay D. Iams, M.D., Ronald J. Wapner, M.D., Deborah Conway, M.D., Mary J. O’Sullivan, M.D., Marshall Carpenter, M.D., Brian Mercer, M.D.,

Susan M. Ramin, M.D., John M. Thorp, M.D., and Alan M. Peaceman, M.D., for the National Institute of Child Health and Human Development Maternal–Fetal Medicine Units Network*

abstract

From Wake Forest University, Winston-Salem, N.C. (P.J.M.); the National Instituteof Child Health and Human Development,Bethesda, Md. (M.K., C.Y.S.); the Biosta-tistics Center, George Washington Univer-sity, Rockville, Md. (E.T.); Wayne State Uni-versity, Detroit (M.P.D.); the University ofTennessee, Memphis (B.S.); the Universityof Chicago, Chicago (A.H.M.); the Univer-sity of Alabama, Birmingham (J.C.H.); theUniversity of Cincinnati, Cincinnati, and Co-lumbia University, New York (M.M.); theUniversity of Utah, Salt Lake City (M.W.V.);the University of Texas Southwestern Med-ical Center, Dallas (K.J.L.); the University ofPittsburgh, Pittsburgh (S.N.C.); Ohio StateUniversity, Columbus (J.D.I.); Thomas Jef-ferson University, Philadelphia (R.J.W.); theUniversity of Texas, San Antonio (D.C.); theUniversity of Miami, Miami (M.J.O.);Brown University, Providence, R.I. (M.C.);Case Western Reserve University, Cleveland(B.M.); the University of Texas, Houston(S.M.R.); the University of North Carolina,Chapel Hill (J.M.T.); and NorthwesternUniversity, Chicago (A.M.P.). Address re-print requests to Dr. Meis at the Depart-ment of Obstetrics and Gynecology, WakeForest University, Medical Center Blvd.,Winston-Salem, NC 27157, or at [email protected].

*Other members of the National Instituteof Child Health and Human DevelopmentMaternal–Fetal Medicine Units Networkare listed in the Appendix.

N Engl J Med 2003;348:2379-85.

Copyright © 2003 Massachusetts Medical Society.

background

Women who have had a spontaneous preterm delivery are at greatly increased risk forpreterm delivery in subsequent pregnancies. The results of several small trials have sug-gested that 17 alpha-hydroxyprogesterone caproate (17P) may reduce the risk of pre-term delivery.

methods

We conducted a double-blind, placebo-controlled trial involving pregnant women witha documented history of spontaneous preterm delivery. Women were enrolled at 19 clin-ical centers at 16 to 20 weeks of gestation and randomly assigned by a central data center,in a 2:1 ratio, to receive either weekly injections of 250 mg of 17P or weekly injections ofan inert oil placebo; injections were continued until delivery or to 36 weeks of gestation.The primary outcome was preterm delivery before 37 weeks of gestation. Analysis wasperformed according to the intention-to-treat principle.

results

Base-line characteristics of the 310 women in the progesterone group and the 153 wom-en in the placebo group were similar. Treatment with 17P significantly reduced the risk ofdelivery at less than 37 weeks of gestation (incidence, 36.3 percent in the progesteronegroup vs. 54.9 percent in the placebo group; relative risk, 0.66 [95 percent confidence in-terval, 0.54 to 0.81]), delivery at less than 35 weeks of gestation (incidence, 20.6 percentvs. 30.7 percent; relative risk, 0.67 [95 percent confidence interval, 0.48 to 0.93]), anddelivery at less than 32 weeks of gestation (11.4 percent vs. 19.6 percent; relative risk,0.58 [95 percent confidence interval, 0.37 to 0.91]). Infants of women treated with 17Phad significantly lower rates of necrotizing enterocolitis, intraventricular hemorrhage,and need for supplemental oxygen.

conclusions

Weekly injections of 17P resulted in a substantial reduction in the rate of recurrent pre-term delivery among women who were at particularly high risk for preterm delivery andreduced the likelihood of several complications in their infants.

The New England Journal of Medicine Downloaded from nejm.org at Bibliothek der MedUni Wien on March 29, 2019. For personal use only. No other uses without permission.

Copyright © 2003 Massachusetts Medical Society. All rights reserved.

n engl j med

348;24

www.nejm.org june

12

,

2003

The

new england journal

of

medicine

2382

ference in the rate of compliance between the twogroups. A total of 231 women (50 percent) reportedat least one adverse effect. The most common sideeffects were local injection-site reactions, includ-ing soreness (in 34.2 percent of the women), swell-ing (in 14.1 percent), itching (in 11.3 percent), andbruising (in 6.7 percent). More women in the pro-gesterone group than in the placebo group hadswelling at the injection site (17.2 percent vs. 7.8percent, P=0.007) or a lump at the injection site(5.5 percent vs. 1.3 percent, P=0.03).

primary outcome and preterm delivery

Outcome data were available for 459 of the 463women (99.1 percent) (Table 2). The frequency ofdelivery before 37 weeks of gestation was 36.3 per-

cent in the progesterone group, as compared with54.9 percent in the placebo group (P<0.001). Deliv-ery before 35 weeks of gestation was also less fre-quent in the progesterone group (20.6 percent vs.30.7 percent, P=0.02). There was a 42 percent re-duction in the rate of delivery before 32 weeks ofgestation in the progesterone group (11.4 percentvs. 19.6 percent, P=0.02). Rates of preterm deliveryin the progesterone group did not differ accordingto the week of gestation at the time of the qualify-ing delivery. Survival analysis showed a significantprolongation of pregnancy with 17P as comparedwith placebo (P=0.01). Because there was an imbal-ance between the progesterone and placebo groupswith regard to the number of previous preterm deliv-eries, we performed an analysis with adjustment forthis variable. The adjusted relative risk of deliverybefore 37 weeks of gestation in the 17P group ascompared with the placebo group was 0.70 (95 per-cent confidence interval, 0.57 to 0.85). There wereno significant differences between the two groupsin the rates of hospital visits for preterm labor, useof tocolytic drugs, corticosteroid use, cesarean de-livery, or chorioamnionitis (Table 2).

More than half the women enrolled were black.The reduction in the rate of preterm delivery with17P among the black women was very similar tothat among nonblack women (Table 2).

The effectiveness of 17P in this study suggeststhat only 5 to 6 women (95 percent confidence in-terval, 3.6 to 11.1) with a level of risk for pretermdelivery similar to that among these women wouldneed to be treated in order to prevent one pretermdelivery before 37 weeks of gestation. Similarly, 12women (95 percent confidence interval, 6.3 to 74.6)with a similar level of risk would need to be treatedin order to prevent one delivery before 32 weeks ofgestation.

Rates of spontaneous miscarriage between 16weeks of gestation and 19 weeks 6 days of gesta-tion, and rates of fetal death after 19 weeks 6 daysof gestation are shown in Tables 2 and 3. There wasa small and nonsignificant increase in the rate ofmiscarriages and stillbirths in the progesteronegroup as compared with the placebo group. Withone exception, all stillbirths occurred before 24weeks of gestation.

outcomes among the infants

There was a significant reduction in the risk of abirth weight of less than 2500 g in the progesteronegroup as compared with the placebo group (relative

* Plus–minus values are means ±SD.† P=0.007.‡ Race was self-assigned by the women.§ The body-mass index is the weight in kilograms divided by the square of the

height in meters.

Table 1. Characteristics of the 463 Women at Randomization.*

Characteristic

Progesterone Group

(N=310)

Placebo Group

(N=153)

Duration of gestation at the time of qualifying delivery — wk

30.6±4.6 31.3±4.2

No. of previous preterm deliveries 1.4±0.7 1.6±0.9†

>1 Previous preterm delivery — no. (%) 86 (27.7) 63 (41.2)

≥1 Previous term deliveries — no. (%) 153 (49.4) 71 (46.4)

Duration of gestation at randomization — wk 18.4±1.4 18.4±1.4

Age — yr 26.0±5.6 26.5±5.4

Race or ethnic group — no. (%)‡Non-Hispanic blackNon-Hispanic whiteHispanicAsianOther

183 (59.0)79 (25.5)43 (13.9)2 (0.6)3 (1.0)

90 (58.8)34 (22.2)26 (17.0)

1 (0.7)2 (1.3)

Marital status — no. (%)Married or living with partnerNever marriedDivorced, widowed, or separated

159 (51.3)119 (38.4)32 (10.3)

71 (46.4)64 (41.8)18 (11.8)

Body-mass index before pregnancy§ 26.9±7.9 26.0±7.0

Yr of education 11.7±2.3 11.9±2.3

Smoking during pregnancy — no. (%) 70 (22.6) 30 (19.6)

Alcohol use during pregnancy — no. (%) 27 (8.7) 10 (6.5)

Substance use during pregnancy — no. (%) 11 (3.5) 4 (2.6)



n engl j med

348;24

www.nejm.org june

12, 2003

progesterone and preterm delivery

2383

risk, 0.66; P=0.003) and a nonsignificant reductionin the risk of a birth weight of less than 1500 g (rel-ative risk, 0.62; P=0.08) (Table 3). Treatment with17P led to significant reductions in the rates of nec-rotizing enterocolitis (P=0.01), need for supple-mental oxygen, and intraventricular hemorrhage ofany grade. However, there was no significant differ-ence between groups in the rate of intraventricularhemorrhage of grade 3 to 4 specifically. The rates ofinfant death, transient tachypnea in the newborn,respiratory distress syndrome, bronchopulmonarydysplasia, need for ventilatory support, retinopathyof prematurity, and patent ductus arteriosus wereslightly but not significantly lower in the progester-one group. Of the 17 neonatal deaths, 16 were dueto complications of prematurity and 1 to intrapar-tum hypoxia subsequent to uterine rupture.

Nine of the infants were found to have congenitalmalformations (2.0 percent in each group). Therewas no consistent pattern to these defects, and noneinvolved genital organs. One infant of a woman inthe progesterone group had torsion of the testiclesin utero, with subsequent infarction.

Treatment with 17P on a weekly basis, beginning at16 to 20 weeks of gestation and continued to deliv-ery or 36 weeks of gestation, significantly reducedthe rate of preterm delivery before 37 weeks, 35weeks, and 32 weeks of gestation among womenat high risk for preterm delivery. The rates of sever-al complications of prematurity were correspond-ingly decreased among the infants of women as-signed to this therapy.

The women enrolled in this study had high ratesof preterm delivery, with more than 50 percent ofthe women who received the placebo injections de-livering before 37 weeks of gestation. This highrate of preterm delivery is most likely related to thehistory of previous preterm deliveries. The earlier ina pregnancy a preterm delivery occurs, the greaterthe chance of preterm delivery in a subsequent preg-nancy.

12

In our study, the mean duration of gesta-tion at the time of the qualifying delivery was 31weeks, and a third of the women enrolled had hadmore than one previous preterm delivery. Therefore,the women in this study had particularly high risk.They were also strongly motivated, and compliancewas excellent.

Preterm delivery has multiple causes.

16

Someevidence suggests that the causes of early preterm

delivery differ from those of later preterm delivery,with earlier preterm deliveries more often being re-lated to infection.

17

Whereas 17P would not be ex-pected to affect an infectious process, in this study,it provided potent protection against early as wellas later preterm delivery. The mechanisms of ac-tion of 17P in prolonging gestation are not entirelyknown. The actions of progesterone on the preg-nant myometrium include relaxation of myometrialsmooth muscle, blocking of the action of oxyto-cin, and inhibition of the formation of gap junc-tions.

18,19

In sheep, goats, and some other mam-mals, a decrease in plasma progesterone and anincrease in circulating estrogen precede the onsetof labor.

20

Although no such alteration in the ratioof plasma estrogen to progesterone precedes theonset of labor in primates, there is evidence thatlocal changes in the progesterone level or the ratioof progesterone to estrogen in the placenta, decid-ua, or fetal membranes may be important in the ini-

discussion

* Data on hospital visit for preterm labor were missing for 1 woman in the pla-cebo group; data on tocolytic therapy were missing for 2 women in the placebo group; and data on corticosteroids for fetal lung maturity were missing for 14 women in the progesterone group and 1 woman in the placebo group. CI de-

notes confidence interval, and NA not applicable.

Table 2. Outcomes of Pregnancy According to Treatment Assignment.*

Outcome

Progesterone Group

(N=306)

Placebo Group

(N=153)

Relative Risk

(95% CI)

no. (%)

Delivery before 37 wk of gestation 111 (36.3) 84 (54.9) 0.66 (0.54–0.81)

Spontaneous 90 (29.4) 69 (45.1) 0.65 (0.51–0.83)

Indicated because of complications

21 (6.9) 15 (9.8) 0.70 (0.37–1.32)

Black women 64 (35.4) 47 (52.2) 0.68 (0.51–0.90)

Nonblack women 47 (37.6) 37 (58.7) 0.64 (0.47–0.87)



Miscarriage at <20 wk of gestation 5 (1.6) 0 NA

Hospital visit for preterm labor 49 (16.0) 21 (13.8) 1.15 (0.72–1.86)

Tocolytic therapy 53 (17.3) 24 (15.9) 1.09 (0.70–1.69)

Corticosteroids for fetal lung maturity

52 (17.8) 30 (19.7) 0.91 (0.60–1.35)

Cesarean delivery 77 (25.2) 41 (26.8) 0.94 (0.68–1.30)

Chorioamnionitis 11 (3.6) 5 (3.3) 1.09 (0.39–3.09)



ALLGEMEINES

KRANKENHAUS




419

Preterm delivery is a leading cause of neonatal mor-bidity and mortality. It is directly responsible for 75% to95% of all neonatal deaths not resulting from lethal con-genital malformations.1,2 Of the survivors, 10% to 15%have significant handicaps.2,3 According to the WorldHealth Organization, a preterm birth is defined as birthbefore 37 completed weeks of gestation.4

In developed countries, the incidence of preterm birthis about 7% to 12% of all deliveries,5,6 and among theseone third occur before 34th week.1 The incidence of pre-term birth in developing countries is higher than in de-veloped countries.1,7,8 In Brazil, preterm birth is a publichealth problem because of the striking social differencesin the population. Because of the high prevalence of

high-risk pregnancies, the incidence of preterm birth atHospital das Clinicas, University of Sao Paulo MedicalSchool, is 22.5%, and half of these resulted from sponta-neous preterm labor. Thus, the prevention of preterm de-livery has become one of the major objectives of perinatalmedicine.

Primary prevention is desirable but not always possi-ble.5,9 The difficulties are due to unawareness of thecause and pathophysiologic mechanisms of pretermbirth, and furthermore, it is not only a medical problem,but also a social and educational problem.

The early detection of pregnant women at high risk forpreterm delivery10-14 could be the best way to prevent preterm birth. Thereby, bed rest, cervical cerclage,15 bac-terial vaginosis treatment, and prophylactic use of pro-gesterone could be one of the managements in thishigh-risk population.

Recent studies have shown that an increase in the num-ber of uterine contractions precedes the onset of pretermlabor,13,16,17 and the frequency of uterine contractions inpregnancies with preterm delivery is higher than inwomen with term and postterm delivery.18

Progesterone is useful in allowing pregnancy to reachits physiologic term because at sufficient levels in the myo-

From the Obstetrics Clinic, University of São Paulo Medical School.Supported by FAPESP.Received for publication January 9, 2002; revised June 12, 2002; ac-cepted September 11, 2002.Reprint requests: Eduardo B. da Fonseca, MD, University of Sao PauloMedical School, Obstetric Clinic, Av 11 de junho, 1006 apto 133, VilaClementino, Sao Paulo–SP, 04041-003, Brazil. E-mail:[email protected]© 2003, Mosby, Inc. All rights reserved.000-9278/2003 $30.00 + 0doi:10.1067/mob.2003.41

Prophylactic administration of progesterone by vaginalsuppository to reduce the incidence of spontaneous preterm birthin women at increased risk: A randomized placebo-controlleddouble-blind study

Eduardo B. da Fonseca, MD, Roberto E. Bittar, PhD, MD, Mario H. B. Carvalho, MD, andMarcelo Zugaib, PhD, MD

Sao Paulo, Brazil

OBJECTIVE: The purpose of this study was to evaluate the effect of prophylactic vaginal progesterone in de-creasing preterm birth rate in a high-risk population.STUDY DESIGN: A randomized, double-blind, placebo-controlled study included 142 high-risk singletonpregnancies. Progesterone (100 mg) or placebo was administered daily by vaginal suppository and all pa-tients underwent uterine contraction monitoring with an external tocodynamometer once a week for 60 min-utes, between 24 and 34 weeks of gestation. Progesterone (n = 72) and placebo (n = 70) groups werecompared for epidemiologic characteristics, uterine contraction frequency, and incidence of preterm birth.Data were compared by χ2 analysis and Fisher exact test.RESULTS: The preterm birth rate was 21.1% (30/142). Differences in uterine activity were found betweenthe progesterone and placebo groups (23.6% vs 54.3%, respectively; P < .05) and in preterm birth betweenprogesterone and placebo (13.8% vs 28.5%, respectively; P < .05). More women were delivered before 34weeks in the placebo group (18.5%) than in the progesterone group (2.7%) (P < .05).CONCLUSION: Prophylactic vaginal progesterone reduced the frequency of uterine contractions and therate of preterm delivery in women at high risk for prematurity. (Am J Obstet Gynecol 2003;188:419-24.)

Key words: Preterm delivery, preterm birth, prevention, progesterone

420 da Fonseca et al February 2003Am J Obstet Gynecol

metrium, it blocks the oxytocin effect of prostaglandinF2α and α-adrenergic stimulation and therefore increasesthe α-adrenergic tocolytic response.6,19 Natural progester-one is free of any disturbing teratogenic, metabolic, or he-modynamic effects. This is not true for certain artificialprogestagens and β-mimetics.1,20

Although some studies demonstrate that natural pro-gesterone is effective in the prevention of preterm deliv-ery and can be administered intramuscularly,21,22 thereare many controversies about their methods. There arefew double-blind studies23,24 that have only used syntheticprogestational agents21,23 with a sufficient number ofwomen. To the best of our knowledge, this is the firststudy that uses natural progestational agents. Therefore,a placebo-controlled clinical trial in asymptomatic high-risk women would be of value.

The objective of this study was to evaluate whether theprophylactic administration of progesterone by vaginalsuppository can reduce the incidence of preterm birth ina high-risk population.

Material and methodsThis study was performed in the Obstetrics Clinic, at

Hospital das Clinicas, University of Sao Paulo MedicalSchool, a tertiary medical center, in Brazil. A consentform was signed after detailed information was given toevery pregnant women. The study was approved by theEthical Commission of this hospital.

Data available at the start of the study showed that thepreterm birth rate in the Obstetrics Clinic, Hospital das

Clinicas, University of Sao Paulo Medical School, was25%.12 Although some studies suggest that the prophy-lactic administration of progesterone in pregnant womenat high risk for preterm birth is associated with a reduc-tion of 60% to 78% in preterm delivery rate,21,23,24 amore realistic assessment of the impact of progesteronemay be a reduction of 50% in the preterm rate.

Therefore, to calculate the sample size, we have pro-posed a reduction of 50% in the preterm birth rate forthe progesterone group (from 25% to 12.5%) and a re-duction of 20% for the placebo group (from 25% to20%).24 A power calculation at the start of the study indi-cated that at least 48 pregnant women would have to beincluded in each group to obtain a study power of 90% ata significance level of .05 (two tailed) to prove the hy-potheses were correct.

Among the women who sought high-risk prenatal care,157 asymptomatic high-risk singleton pregnant women forpreterm delivery were followed up from February 2, 1996,to March 30, 2001. Patients were allocated to progesteroneor placebo according to a randomized number table. Thenumbers corresponded to sealed envelopes that indicatedif drug A or drug B should be used. Numbers were givenconsecutively. Treatment assignment was blinded until thedelivery of the last pregnant woman. Both the patients andthe staff who were recording the study findings wereblinded to the study medication allocation.

Fifteen (9.5%) patients were lost to follow-up or with-drew from the study. Nine (11.1 %) of these were in theprogesterone group and 5 (6.5 %) in the placebo group,resulting in 72 assigned to the progesterone group and70 to the placebo group. None of the patients in eitherstudy group had a multiple pregnancy.

Women at high risk for preterm delivery were consid-ered to be those in the presence of at least one previousspontaneous preterm birth, prophylactic cervical cer-clage, and uterine malformation. Gestational age at aprior preterm birth for the progesterone and placebogroups was 33.3 (±2.7) and 33.4 (±2.6) weeks. We did notobserve a significant difference in the gestational age ofprevious preterm birth, uterine malformation, and cervixcerclage in these two groups. Multiple gestation and fetalmalformations were excluded.

Women allergic to progesterone (n = 1), who missedfollow-up (n = 1), those with preterm rupture of mem-

Table I. Exclusion cause in two study groups

Placebo ProgesteroneExclusion cause (n = 76) (n = 81) P value

PROM 4 (5.2%) 6 (7.4%) NSLost follow-up 1 (1.3%) 0 NSTherapeutic preterm delivery 1 (1.3%) 2 (2.4%) NSAllergic process 0 1 (1.2%)

PROM, Preterm rupture of membranes; NS, not significant.

Table II. Characteristics of women at randomization

Placebo Progesterone(n = 70) (n = 72)

Age (y)* 26.8 27.6Ethnicity*

White 71.4% 68.0%Nonwhite 28.6% 32.0%

Parity (>1 delivery)* 97.1% 90.2%Risk factor*

Previous preterm delivery 97.2% 90.3%Uterine malformation 1.4% 5.6%Incompetent cervix 1.4% 4.1%

Gestational age at intake (wk)* 25.2 26.5

*Not significant.422 da Fonseca et al February 2003

Am J Obstet Gynecol

stetric history. There was no significant difference be-tween gestational age at study admission and vaginal in-fection. Socioeconomic status, estimated by theeducational level as well as ethnicity, was similar in bothgroups.

Twenty-two women in the placebo group (31.4%) and14 in the progesterone group (19.4%) were admitted forpreterm labor. However, this difference was not signifi-cant. The use of β-mimetic drugs in the management ofpreterm labor demonstrated a significant benefit in theprogesterone group (P = .031). In the progesteronegroup, 85.7% of pregnant women had their delivery de-layed for more than 72 hours, whereas in the placebogroup this was observed in only 36.4% of the patients.Twelve of the 22 pregnant women in the placebo group(54.5%) and 10 of the 14 pregnant women (71.4%) inthe progesterone group had a second episode of pretermlabor, with an interval time of 3.9 ± 3.2 days and 5.7 ± 2.3days in the placebo and progesterone groups, respec-tively (P = .02).

The average gestational age for those who had pretermbirth was 33.5 ± 2.4 weeks in the progesterone group and32.0 ± 0.7 weeks in the placebo group. Because pretermbirth before 34 weeks is associated with the worst preg-

nancy outcome, we were especially interested in decreas-ing preterm birth incidence in this period. In Table III, itcan be seen that more women were delivered before 34weeks in the placebo group (18.6%) than in the proges-terone group (2.8%). Figs 1 and 2 show the frequency ofpreterm delivery before 34 weeks. When the difference inthe frequency of preterm birth in the progesterone(2.8%) and placebo (18.8 %) groups was compared, a sta-tistically significant difference was observed (P = .002).

When survival analysis was used to establish the rela-tionship between prophylactic vaginal progesterone ad-ministration and preterm birth, a lower gestational age atdelivery correlated with the placebo group (mean 36 ±3.3 weeks, range 29–41 vs mean 37 ± 2.8 weeks, range28–41; P = .029). The probability of undelivered patientsat 34 weeks of gestation was higher in the progesteronegroup than in the placebo group (97.2% vs 81.4%; P =.029) (Fig 3).

Mean contraction frequency for each gestational weekstudied was significantly greater for the placebo groupthan the progesterone group (Table IV). We calculatedthe maximum number of contractions per hour for eachweek between 28 and 34 weeks’ gestation. The frequencyof contractions was inferior in the group treated with pro-gesterone than in the placebo group (P < .004) (Fig 4). Asshown in Table V, the frequency of uterine contractionsof more than four contractions per hour was more fre-quently found in the placebo group than in the proges-terone group (54.3% vs 23.6%, respectively; P = .0001).

Table IV. Mean contraction frequency for each gesta-tional week between placebo and progesterone groups

Placebo Progesterone

Gestational age (wk) Mean ± SD Mean ± SD P value

28 4.0 ± 3.0 1.0 ± 0.6 .0000129 4.0 ± 2.1 1.0 ± 0.9 .0000130 6.2 ± 3.0 2.8 ± 2.7 .0000131 5.1 ± 2.5 3.2 ± 2.0 .000132 6.5 ± 3.1 2.5 ± 2.5 .0133 7.0 ± 4.2 2.8 ± 2.4 .000134 6.5 ± 3.1 3.5 ± 2.0 .0001

Fig 3. Cumulative percentage of undelivered patients per week,by placebo and progesterone group. Log-rank χ2 = 5.33, P = .029.

Fig 4. Mean contraction frequency based on 1 hour of monitor-ing per week, by placebo and progesterone group and by gesta-tional age (28, 29, 30, 31, 32, 33, and 34 weeks’ gestation). Opensquares, Placebo group; solid circle, progesterone group. F = 9.5, P = .004.

Table V. Frequency of uterine contraction

Placebo Progesterone Contraction (n = 70) (n = 72) P value

<4 32 (45.7%) 55 (76.4%) .00014–5 12 (17.1%) 3 (4.1%) .0118≥6 26 (37.2%) 14 (19.4%) .0190

Fonseca et al., 2003

Volume 188, Number 2 da Fonseca et al 421Am J Obstet Gynecol

branes (PROM) (n = 10), and those having a therapeuticpremature delivery (n = 3) were excluded from the study.One hundred forty-two women completed the study, andthere was no statistically significant difference for the ex-clusion cause in both groups (Table I).

Gestational age was calculated on the basis of the lastmenstrual period and ultrasonography up to 12 weeks orby two concordant scans between 12 and 20 weeks.

At the first prenatal visit, a microscopic examinationand culture of cervicovaginal secretions for Trichomonasvaginalis, Candida sp, Neisseria gonorrhoeae, Chlamydia tra-chomatis, Gardnerella vaginalis, and group B Streptococcuswere carried out for all patients. Women with positivecultures were treated with specific antibiotics, and re-peat cultures were performed to confirm the efficacy oftreatment.

All pregnant women were submitted to uterine con-traction monitoring by an external tocodynamometeronce a week for 60 minutes from 24 to 34 weeks of gesta-tion. We performed uterine monitoring between 8 and10 AM using a Hewlett Packard tocodynamometer 50A se-ries (Hewlett Packard, Houston, Tex) while women werein semi-Fowler position. We determined the frequency ofcontractions and compared mean values of both studygroups. A positive test was considered when there werefour or more contractions per hour before the 30th weekof gestation and from 30 weeks onward, 6 or more con-tractions per hour.12,21,23

Preterm labor was defined as two or more regular uter-ine contractions every 10 minutes, recorded by externaltocodynamometer, associated with cervical changes, rep-resented by a dilatation of more than 2 cm, or the pres-ence of progressive dilatation or effacement of the cervix.Women in preterm labor were treated in the hospital withintravenous tocolytic therapy. A preterm delivery was de-fined as birth before 37 weeks of pregnancy.

Both groups of pregnant women were randomly se-lected to receive the vaginal progesterone suppository(100 mg) or an identical-looking placebo. The supposito-ries were identical in appearance and thick. They wereapplied every night from 24 to 34 weeks of gestation. Pa-tients had a thorough explanation of how to use the sup-positories, including an orientation picture. Themedication and the placebo were supplied by manipula-tion pharmacy at Hospital das Clínicas, University of SaoPaulo. Patient treatment was only unblinded after the de-livery of the last pregnant women.

The clinical relevance of the prophylactic use of pro-gesterone was determined as it correlated with the evolu-tion of pregnancy to preterm delivery. Statistical analysiswas performed with EPI-INFO 2000 1.0 (Centers for Dis-ease Control and Prevention, Atlanta, Ga) and STATA 7.0(USA) (Stata, College Station, Tex). The χ2 tests or Fisherexact test were used for categoric variables. The two-tailed Student t test was used for continuous variables andthe Wilcoxon rank sum test was used for interval vari-ables. Kaplan-Meier survival analysis was used to deter-mine the relationship between the administration ofprophylactic vaginal progesterone and preterm birth.The log-rank χ2 test was used to compare the differencesin the generated survival curves. A P value of .05 was con-sidered significant.

ResultsOf 142 cases, there were 30 preterm births (21.1%).

The incidence of preterm delivery in the progesteronegroup was 13.8% (10/72) and 28.5% (20/70) in theplacebo group. When comparing these two groups, weobserved a statistically significant difference in the pre-term delivery rate (P = .03).

As shown in Table II, the two groups were found similarin regard to age, risk factors for preterm delivery, and ob-

Fig 1. Incidence of preterm delivery before 34th week in naturalprogesterone group.

Fig 2. Percentage of preterm delivery before 34th week inplacebo group.

Table III. Incidence of preterm delivery

Placebo Progesterone(n = 70) (n = 72) P value

<37 wk 20 (28.5%) 10 (13.8%) .0334 wk 13 (18.6%) 2 (2.8%) .002Admission for threatened preterm labor 22 (31.4%) 14 (19.4%) NS

NS, Not significant.

ALLGEMEINES

KRANKENHAUS




Hassan et al., 2000

ALLGEMEINES

KRANKENHAUS




Fonseca et al., 2007

original article

T h e n e w e ng l a nd j o u r na l o f m e dic i n e

n engl j med 357;5 www.nejm.org august 2, 2007462

Progesterone and the Risk of Preterm Birth among Women with a Short CervixEduardo B. Fonseca, M.D., Ebru Celik, M.D., Mauro Parra, M.D.,

Mandeep Singh, M.D., and Kypros H. Nicolaides, M.D., for the Fetal Medicine Foundation Second Trimester Screening Group*

From the Harris Birthright Research Cen-tre for Fetal Medicine, King’s College Hos-pital, London. Address reprint requests to Dr. Nicolaides at the Harris Birthright Research Centre for Fetal Medicine, King’s College Hospital Medical School, Den-mark Hill, London SE5 8RX, United King-dom, or at [email protected].

*The other members of the Fetal Medicine Foundation Second Trimester Screening Group are listed in the Appendix.

N Engl J Med 2007;357:462-9.Copyright © 2007 Massachusetts Medical Society.

A BS TR AC T

BACKGROUNDPrevious randomized trials have shown that progesterone administration in women who previously delivered prematurely reduces the risk of recurrent premature delivery. Asymptomatic women found at midgestation to have a short cervix are at greatly increased risk for spontaneous early preterm delivery, and it is unknown whether progesterone reduces this risk in such women.

METHODSCervical length was measured by transvaginal ultrasonography at a median of 22 weeks of gestation (range, 20 to 25) in 24,620 pregnant women seen for routine prenatal care. Cervical length was 15 mm or less in 413 of the women (1.7%), and 250 (60.5%) of these 413 women were randomly assigned to receive vaginal proges-terone (200 mg each night) or placebo from 24 to 34 weeks of gestation. The pri-mary outcome was spontaneous delivery before 34 weeks.

RESULTSSpontaneous delivery before 34 weeks of gestation was less frequent in the proges-terone group than in the placebo group (19.2% vs. 34.4%; relative risk, 0.56; 95% confidence interval [CI], 0.36 to 0.86). Progesterone was associated with a nonsig-nificant reduction in neonatal morbidity (8.1% vs. 13.8%; relative risk, 0.59; 95% CI, 0.26 to 1.25; P = 0.17). There were no serious adverse events associated with the use of progesterone.

CONCLUSIONSIn women with a short cervix, treatment with progesterone reduces the rate of spon-taneous early preterm delivery. (ClinicalTrials.gov number, NCT00422526.)



Progesterone for Women with a Short Cervix

n engl j med 357;5 www.nejm.org august 2, 2007 467

the progesterone group (34 of 109 [31.2%] vs. 20 of 112 [17.9%]; relative risk, 0.57; 95% CI, 0.35 to 0.93; P = 0.03). Among women with singleton pregnancies, the incidence of spontaneous pre-term birth was significantly higher in the placebo group than in the progesterone group (36 of 112 [32.1%] vs. 20 of 114 [17.5%]; relative risk, 0.54; 95% CI, 0.34 to 0.88; P = 0.02).

There were no significant differences between the two groups in the secondary outcomes (Ta-ble 2). There were no important adverse events or side effects in either group. None of the women reported any increase in the frequency or severity of general or local side effects, such as sleepiness, fatigue, headaches, or genital irritation, or any new symptoms after the onset of treatment.

Data on pregnancy outcome were also ob-tained from 23,795 (96.6%) of the 24,620 women in whom cervical length was originally measured; spontaneous delivery before 34 weeks occurred in 489 (2.1%). The cervical length was 15 mm or less in 408 women (1.7%), of whom 126 (30.9%) delivered preterm, accounting for 25.8% of the early deliveries. The cervical length was 16 to 25 mm in 1975 women (8.3%), of whom 100 (5.1%) delivered preterm, accounting for 20.4% of the early deliveries.

Discussion

The results of this randomized trial demonstrate that in women with a short cervix, the daily vag-inal administration of 200 mg of progesterone from 24 to 34 weeks of gestation significantly reduces the rate of spontaneous preterm delivery. There was no significant reduction in perinatal mortality or neonatal morbidity. However, the trial was not designed with sufficient power to address these end points.

Our multicenter screening study, involving close to 25,000 pregnancies, confirms that trans-vaginal ultrasonographic measurement of cervi-cal length at 22 weeks of gestation identifies a subgroup of about 1.5% of the female population at particularly high risk for early preterm deliv-ery. In the control group of women with cervical lengths of 15 mm or less, the incidence of spon-taneous early preterm delivery was 34%. This is much higher than the overall female population rate of about 2% in the United Kingdom.17

In the small number of twin pregnancies in-cluded in our study, a nonsignificant reduction

in preterm delivery was associated with proges-terone treatment. In a larger study published else-where in this issue of the Journal that examined twin pregnancies specifically,18 the intramuscu-lar administration of 17 alpha-hydroxyprogester-one caproate (17P) did not reduce the incidence of preterm birth.

We used 200 mg of progesterone, in contrast to the 100-mg dose used in a randomized trial of women with a history of preterm birth.5 We chose this high dose because we considered patients with a very short cervix to be at particularly high risk for preterm delivery,8-10 although it is un-known whether there is a dose–response relation-ship between progesterone and the reduction in risk of preterm delivery. We chose vaginal micron-ized natural progesterone, rather than intramus-cular synthetic 17P. Micronized progesterone can be administered either orally or vaginally, but the latter route is preferable because of enhanced

100

*#

*"�)

+��

��'�

�$)�

��%

��%$

) $*�

��'

��$�

$� �

(

80

90

70

60

00 160 170 180 190 200 210 220 230 240

Label 1

Label 2

��()�) %$�"��-(�

�%��)�� (!��%'�&%$)�$�%*(�� ')�

ProgesteronePlacebo

125125

125121

122119

118115

114109

112105

10798

10387

9980

AUTHOR:

FIGURE:

JOB:

4-CH/T

RETAKE

��

ICM

CASE

EMail LineH/TCombo

Revised

�� *'��(��$�'��'�,$��$��)-&��(��$�'�(�)�

�"��(��!��'��*""-�

REG F

Enon

1st2nd3rd

Fonseca (Nicolaides)

2 of 3

07-12-07

ARTIST: ts

35702 ISSUE:

22p3

Progesterone

Placebo

P=0.02

Figure 2. Kaplan–Meier Plot of the Probability of Continued Pregnancy without Delivery among Patients Receiving Vaginal Progesterone as Compared with Placebo.

Progesterone reduces the risk of spontaneous delivery before 34 weeks by 44.2% (hazard ratio for progesterone, 0.57; 95% CI, 0.35 to 0.92; P = 0.02). P = 0.49 for the test of the proportional-hazards assumption.



T h e n e w e ng l a nd j o u r na l o f m e dic i n e

n engl j med 357;5 www.nejm.org august 2, 2007466

women (1.7%). Two hundred fifty women with a short cervix (60.5%), including 226 with single-ton and 24 with twin pregnancies, agreed to par-ticipate in the trial.

There were no significant differences in base-line characteristics between the placebo and the progesterone groups (Table 1). The rate of the primary outcome — spontaneous birth before 34 weeks of gestation — was 19.2% in the pro-gesterone group and 34.4% in the placebo group (relative risk, 0.56; 95% confidence interval [CI], 0.36 to 0.86) (Table 2). Four women (two in each group) had medically indicated preterm delivery. The proportional reduction over time in the in-cidence of all deliveries before 34 weeks was simi-lar in the progesterone and the control groups. The risk of spontaneous preterm birth in the two groups was assessed using Kaplan–Meier analysis

(Fig. 2). The cumulative percentage of patients who did not give birth spontaneously before 34 weeks was significantly higher in the progester-one group than in the placebo group (hazard ratio, 0.57; 95% CI, 0.35 to 0.92; P = 0.02). Multi-variable analysis demonstrated that adjustment for maternal characteristics at the time of ran-domization did not attenuate the apparent pro-tective effect of progesterone (Table 2).

The relative risk of spontaneous preterm birth before 34 weeks of gestation did not vary signifi-cantly according to maternal age, body-mass in-dex, race, obstetrical history, whether the preg-nancy was singleton or twin, or cervical length at the time of randomization (Fig. 3). Among wom-en without a history of delivery before 34 weeks, the incidence of spontaneous preterm birth was significantly higher in the placebo group than in

Table 2. Outcomes According to Study Group.*

OutcomeProgesterone

Group†PlaceboGroup‡

Relative Risk(95% CI) P Value

AdjustedRelative Risk

(95% CI) P Value

no. (%)

Maternal

Spontaneous delivery at <34 wk 24 (19.2) 43 (34.4) 0.56 (0.36–0.86) 0.007 0.56 (0.32–0.91) 0.02

Any delivery at <34 wk 26 (20.8) 45 (36.0) 0.58 (0.38–0.87) 0.008 0.60 (0.35–0.94) 0.02

Perinatal

Fetal death 1 (0.7) 1 (0.7) 0.98

Neonatal death 2 (1.5) 7 (5.1) 0.29 (0.06–1.42) 0.13 0.34 (0.06–1.81) 0.22

Birth weight <2500 g 56 (41.2) 59 (42.8) 0.96 (0.69–1.26) 0.81 0.97 (0.68–1.29) 0.85

Birth weight <1500 g 18 (13.2) 27 (19.6) 0.68 (0.36–1.21) 0.20 0.74 (0.36–1.37) 0.35

Composite adverse outcomes 11 (8.1) 19 (13.8) 0.59 (0.26–1.25) 0.17 0.57 (0.23–1.31) 0.19

Intraventricular hemorrhage§ 1 (0.7) 2 (1.4) 0.51 (0.05–5.30) 0.58 0.33 (0.01–8.84) 0.52

Respiratory distress syndrome 11 (8.1) 19 (13.8) 0.59 (0.26–1.25) 0.17 0.57 (0.23–1.31) 0.19

Retinopathy of prematurity 2 (1.5) 0

Necrotizing entercolitis 0 1 (0.7)

Composite therapy 34 (25.0) 45 (32.6) 0.77 (0.48–1.15) 0.21 0.75 (0.44–1.16) 0.20

Neonatal intensive care 33 (24.3) 42 (30.4) 0.80 (0.49–1.21) 0.30 0.80 (0.47–1.24) 0.34

Ventilation 16 (11.8) 25 (18.1) 0.65 (0.33–1.21) 0.18 0.64 (0.30–1.25) 0.20

Phototherapy 16 (11.8) 14 (10.1) 1.16 (0.56–2.25) 0.68 1.09 (0.50–2.19) 0.82

Treatment for sepsis 3 (2.2) 11 (8.0) 0.28 (0.07–1.01) 0.05 0.29 (0.07–1.10) 0.07

Blood transfusion 4 (2.9) 5 (3.6) 0.81 (0.22–2.86) 0.75 0.79 (0.19–3.10) 0.74

* For perinatal outcomes, the relative risks, 95% confidence intervals, and P values were estimated by logistic regression clustered on maternal identifiers to account for nonindependence between twin pairs. Relative risks were adjusted for maternal age, body-mass index, smoking status, race, history of preterm birth, and cervical length at the time of ran-domization.

† There were 125 pregnancies and 136 infants in the progesterone group.‡ There were 125 pregnancies and 138 infants in the placebo group.§ Intraventricular hemorrhage was grade 2 in all infants.



ALLGEMEINES

KRANKENHAUS




Romero et al., 2012

Figure 3.Effect of vaginal progesterone on preterm birth before 33 weeks of gestation

ROMERO et al. Page 26

Am J Obstet Gynecol. Author manuscript; available in PMC 2012 September 10.

NIH

-PA Author Manuscript

NIH


NIH


Geburt < SSW 33

ALLGEMEINES

KRANKENHAUS




Romero et al., 2012

ALLGEMEINES

KRANKENHAUS




Wirkung von vaginalem Progesteron auf Patienten mitvorangegangener Frühgeburt und kurzer Zervix

Romero et al., 2012

ALLGEMEINES

KRANKENHAUS




Wirkung von vaginalem Progesteron auf Patientenmit vorangegangener Frühgeburt und kurzer Zervix - kindliches Outcome

Romero et al., 2012

ALLGEMEINES

KRANKENHAUS




Effekt von Progesteron auf Zwillinge mit verkürzter Cervix

Romero et al., 2012

ALLGEMEINES

KRANKENHAUS




Effekt von Progesteron auf Zwillinge mit verkürzter Cervix

Romero et al., 2012

* Defined as the occurrence of any of the following events: respiratory distress syndrome, intraventricular hemorrhage, necrotizing enterocolitis, proven neonatal sepsis, or neonatal death.

ALLGEMEINES

KRANKENHAUS




Norman et al., 2016

Articles

2106 www.thelancet.com Vol 387 May 21, 2016

Vaginal progesterone prophylaxis for preterm birth (the OPPTIMUM study): a multicentre, randomised, double-blind trialJane Elizabeth Norman, Neil Marlow, Claudia-Martina Messow, Andrew Shennan, Phillip R Bennett, Steven Thornton, Stephen C Robson, Alex McConnachie, Stavros Petrou, Neil J Sebire, Tina Lavender, Sonia Whyte, John Norrie, for the OPPTIMUM study group

SummaryBackground Progesterone administration has been shown to reduce the risk of preterm birth and neonatal morbidity in women at high risk, but there is uncertainty about longer term effects on the child.

Methods We did a double-blind, randomised, placebo-controlled trial of vaginal progesterone, 200 mg daily taken from 22–24 to 34 weeks of gestation, on pregnancy and infant outcomes in women at risk of preterm birth (because of previous spontaneous birth at ≤34 weeks and 0 days of gestation, or a cervical length ≤25 mm, or because of a positive fetal fibronectin test combined with other clinical risk factors for preterm birth [any one of a history in a previous pregnancy of preterm birth, second trimester loss, preterm premature fetal membrane rupture, or a history of a cervical procedure to treat abnormal smears]). The objective of the study was to determine whether vaginal progesterone prophylaxis given to reduce the risk of preterm birth affects neonatal and childhood outcomes. We defined three primary outcomes: fetal death or birth before 34 weeks and 0 days gestation (obstetric), a composite of death, brain injury, or bronchopulmonary dysplasia (neonatal), and a standardised cognitive score at 2 years of age (childhood), imputing values for deaths. Randomisation was done through a web portal, with participants, investigators, and others involved in giving the intervention, assessing outcomes, or analysing data masked to treatment allocation until the end of the study. Analysis was by intention to treat. This trial is registered at ISRCTN.com, number ISRCTN14568373.

Findings Between Feb 2, 2009, and April 12, 2013, we randomly assigned 1228 women to the placebo group (n=610) and the progesterone group (n=618). In the placebo group, data from 597, 587, and 439 women or babies were available for analysis of obstetric, neonatal, and childhood outcomes, respectively; in the progesterone group the corresponding numbers were 600, 589, and 430. After correction for multiple outcomes, progesterone had no significant effect on the primary obstetric outcome (odds ratio adjusted for multiple comparisons [OR] 0·86, 95% CI 0·61–1·22) or neonatal outcome (OR 0·72, 0·44–1·17), nor on the childhood outcome (cognitive score, progesterone group vs placebo group, 97·3 [SD 17·9] vs 97·7 [17·5]; difference in means –0·48, 95% CI –2·77 to 1·81). Maternal or child serious adverse events were reported in 70 (11%) of 610 patients in the placebo group and 59 (10%) of 616 patients in the progesterone group (p=0·27).

Interpretation Vaginal progesterone was not associated with reduced risk of preterm birth or composite neonatal adverse outcomes, and had no long-term benefit or harm on outcomes in children at 2 years of age.

Funding Efficacy and Mechanism Evaluation (EME) Programme, a Medical Research Council (MRC) and National Institute for Health Research (NIHR) partnership. The EME Programme is funded by the MRC and NIHR, with contributions from the Chief Scientist Office in Scotland and National Institute for Social Care and Research in Wales.

Copyright © Norman et al. Open Access article distributed under the terms of CC BY.

IntroductionSeveral studies have assessed either vaginal progesterone or intramuscular 17α-hydroxyprogesterone caproate for the prevention of preterm birth in asymptomatic women with singleton pregnancy at high risk of preterm birth. An individual patient data meta-analysis of women with a short cervix showed that vaginal progesterone reduced the risk of preterm birth before 33 weeks (relative risk [RR] 0·58, 95% CI 0·42–0·80) and reduced a composite of neonatal mortality and morbidity (RR 0·57, 0·40–0·81).1 Although there is debate whether vaginal

and intramuscular therapies have similar mechanisms or efficacy, the Cochrane Library meta-analysis groups the two treatments together, but reports separately for different maternal risk groups.2 Reduced risk of preterm birth before 34 weeks was shown in women with a short cervix (RR 0·64, 95% CI 0·45–0·90), without effect on perinatal mortality or neonatal death (perinatal mortality RR 0·74, 0·42–1·29; neonatal death RR 0·55, 0·26–1·13).2 By contrast, in women with previous preterm birth, progestogens reduced the incidence of preterm birth (RR 0·31, 95% CI 0·14–0·69), perinatal

Lancet 2016; 387: 2106–16

Published Online February 23, 2016

http://dx.doi.org/10.1016/S0140-6736(16)00350-0

This online publication has been corrected. The corrected version first appeared at thelancet.com

on January 17, 2019

See Comment page 2066

Tommy’s Centre for Maternal and Fetal Health, MRC Centre

for Maternal and Fetal Health, University of Edinburgh,

Edinburgh, UK (Prof J E Norman MD,

S Whyte MSc); University College London, London, UK

(Prof N Marlow DM, Prof N J Sebire MD); Robertson

Centre for Biostatistics, Institute of Health and

Wellbeing, University of Glasgow, Glasgow, UK

(C-M Messow PhD, A McConnachie PhD); Women’s

Health Academic Centre, King’s College London, London, UK

(Prof A Shennan MD); Imperial College London, London, UK (Prof P R Bennett MD); Queen

Mary University of London, London, UK

(Prof S Thornton DM); Medical School, University of Newcastle,

Newcastle, UK (Prof S C Robson MD); Division

of Health Sciences, Warwick Medical School, University of

Warwick, Coventry, UK (Prof S Petrou PhD); University

of Manchester School of Nursing, University of

Manchester, Manchester, UK (Prof T Lavender PhD); and

Centre for Healthcare Randomised Trials, Health

Services Research Unit, University of Aberdeen,

Aberdeen, UK (Prof J Norrie MSc)

Articles

www.thelancet.com Vol 387 May 21, 2016 2111

Placebo group Progesterone group Unadjusted odds ratio (95% CI) or difference in means (95% CI)

p value (unadjusted)

Adjusted odds ratio (95% CI)* or difference in means (95% CI)

p value (adjusted*)

Fetal death or delivery <34 weeks of gestation 108/597 (18%) 96/600 (16%) 0·86 (0·64 to 1·17) 0·34 0·86 (0·61 to 1·22) 0·67

Neonatal morbidity or death 62/587 (11%) 46/589 (8%) 0·72 (0·48 to 0·94) 0·10 0·72 (0·44 to 1·17) 0·31

Cognitive composite score at 2 years†‡ 97·7 (17·5) 97·3 (17·9) –0·48 (–2·77 to 1·81)§ 0·68 –0·48 (–2·77 to 1·81)§ 0·68

Components of the obstetric outcome

Fetal death 7/597 (1%) 8/600 (1%) 1·14 (0·41 to 3·17) 0·8 ·· ··

Liveborn delivery before 34 weeks 101/590 (17%) 88/592 (15%) 0·85 (0·62 to 1·15) 0·29 ·· ··

Components of the neonatal outcome

Neonatal death 8/597 (1%) 9/600 (2%) 1·14 (0·44 to 2·98) 0·79¶ ·· ··

Bronchopulmonary dysplasia|| 18/574 (3%) 17/580 (3%) 0·94 (0·49 to 1·78) 0·84 ·· ··

Brain injury on ultrasound scan** 34/574 (6%) 18/584 (3%) 0·50 (0·31 to 0·84) 0·008 ·· ··

Binary outcomes are n/N (%) and continuous outcomes are mean (SD). *CI for odds ratio (OR) and p value adjusted for multiple primary outcomes using Bonferroni-Holm method. †Median weeks of age at assessment: 111·6 weeks (IQR 104·6–122·2) in the placebo group and 110·4 weeks (104·0–121·5) in the progesterone group. ‡Sample size of 439 in the placebo group and 430 in the progesterone group and includes imputations for deaths. §Difference in means (95% CI). ¶Unadjusted for previous pregnancy of at least 14 weeks because of small sample size. ||Bronchopulmonary dysplasia defined as need for at least 30% oxygen to maintain oxygen saturation above 92% or positive pressure (positive pressure ventilation or nasal continuous positive airway pressure) at 36 weeks postmenstrual age or discharge, whichever comes first. **Brain injury on ultrasound scan defined as any intraventricular haemorrhage (excludes subependymal haemorrhages), parenchymal cystic or haemorrhagic lesion, or persistent ventriculomegaly (ventricular index >97th percentile); the components of the brain scan abnormalities were: intraventricular haemorrhage 13 (3%) of 383 patients and seven (2%) of 357 patients, parenchymal cystic or haemorrhagic lesion 23 (6%) of 382 and eight (2%) of 357, and persistent ventriculomegaly (>97th percentile) eight (2%) of 372 and three (1%) of 349 in the placebo group and the progesterone group, respectively.

Table 2: Primary outcomes and their components for women entered into the treatment phase of the OPPTIMUM study and their babies

Treatment effect pinteraction

OR or mean difference (95% CI); p value

N OR or mean difference (95% CI); p value

N

Fibronectin status Negative Negative Positive Positive

Obstetric outcome 0·88 (0·58 to 1·33); 0·542 859 0·91 (0·57 to 1·46); 0·707 338 0·91

Neonatal outcome 0·78 (0·46 to 1·33); 0·357 847 0·69 (0·37 to 1·30); 0·254 329 0·79

Childhood outcome –0·63* (–3·28 to 2·03); 0·644 628 –1·09* (–5·41 to 3·23); 0·612 241 0·86

Cervical length at baseline >25 mm >25 mm ≤25 mm ≤25 mm




Cervical length at baseline >15 mm >15 mm ≤15 mm ≤15 mm




Chorioamnionitis No No Yes Yes




History of spontaneous preterm birth No No Yes Yes




History of any preterm birth No No Yes Yes




Logistic or linear mixed effects regression models predicting outcome from treatment, subgroup and the interaction of treatment with the subgroup variable, adjusting for previous pregnancy of at least 14 weeks and a random effect for study centre. *Mean difference.

Table 3: Prespecified subgroup analyses based on baseline risk factors in women entered into the treatment phase of the OPPTIMUM study

ALLGEMEINES

KRANKENHAUS




Romero et al., 2018

VAGINAL PROGESTERONE FOR PREVENTING PRETERM BIRTH AND ADVERSE PERINATAL OUTCOMES IN SINGLETON GESTATIONS WITH A SHORT CERVIX: A META-ANALYSIS OF INDIVIDUAL PATIENT DATA

Roberto Romero, MD, DMedSci1,2,3,4, Agustin Conde-Agudelo, MD, MPH, PhD1,5, Eduardo Da Fonseca, MD6, John M. O’Brien, MD7, Elcin Cetingoz, MD8, George W. Creasy, MD9, Sonia S. Hassan, MD1,5, and Kypros H. Nicolaides, MD10

1Perinatology Research Branch, Program for Perinatal Research and Obstetrics, Division of Intramural Research, Eunice Kennedy Shriver National Institute of Child Health and Human Development, National Institutes of Health, Department of Health and Human Services, Bethesda, MD and Detroit, MI, USA2Department of Obstetrics and Gynecology, University of Michigan, Ann Arbor, MI, USA3Department of Epidemiology and Biostatistics, Michigan State University, East Lansing, MI, USA4Center for Molecular Medicine and Genetics, Wayne State University, Detroit, MI, USA5Department of Obstetrics and Gynecology, Wayne State University School of Medicine, Detroit, MI, USA6Departamento de Obstetrícia e Ginecologia, Hospital do Servidor Publico Estadual “Francisco Morato de Oliveira” and School of Medicine, University of São Paulo, São Paulo, Brazil7Department of Obstetrics and Gynecology, University of Kentucky, Lexington, KY, USA8Department of Obstetrics and Gynecology, Zeynep Kamil Women and Children Diseases Education and Research Hospital, Uskudar, Istanbul, Turkey9Center for Biomedical Research, Population Council, New York, NY10Harris Birthright Research Centre for Fetal Medicine, King’s College Hospital, London, UK

Corresponding author Dr. Roberto Romero, Perinatology Research Branch, Intramural Division, NICHD/NIH/DHHS, Hutzel Women’s Hospital, Box # 4, 3990 John R, Detroit, MI 48201, Telephone: +1 313 993 2700, Fax: +1 313 993 2694, [email protected]. Disclosure: RR, AC-A, EDF, EC, SSH, and KHN declare no conflict of interest. JMO’B was involved in studies of progesterone gel treatment for preterm birth prevention sponsored by a maker of progesterone gel. He served on advisory boards and as a consultant for Watson Pharmaceuticals, a company with a financial interest in marketing vaginal progesterone gel for preterm birth prevention; he and others are listed in a patent on the use of progesterone compounds to prevent preterm birth (USA Patent Number 7884093: progesterone for the treatment and prevention of spontaneous preterm birth). He has received no royalty payments. GWC was an Employee of Columbia Laboratories, Inc. when the previous meta-analysis of individual patient data was conducted in 2011.Professor Jane Norman has no conflict of interest in relation with our meta-analysis of individual patient data.Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

HHS Public AccessAuthor manuscriptAm J Obstet Gynecol. Author manuscript; available in PMC 2019 February 01.

Published in final edited form as:Am J Obstet Gynecol. 2018 February ; 218(2): 161–180. doi:10.1016/j.ajog.2017.11.576.

Author Manuscript

Author Manuscript

Author Manuscript

Author Manuscript

Figure 3. Effect of vaginal progesterone on preterm birth <33 weeks of gestation

Romero et al. Page 29

Am J Obstet Gynecol. Author manuscript; available in PMC 2019 February 01.

Author Manuscript

Author Manuscript

Author Manuscript

Author Manuscript

Geburt <SSW33

ALLGEMEINES

KRANKENHAUS




Wirkt 17 Hydoxyprogesteron bei Frauen mitkurzer Zervix ohne vorangegangene Frühgeburt

Grobman et al., 2012

Figure 2.Survival curve illustrating the proportion of participants remaining pregnant afterrandomization.

Grobman et al. Page 11

Am J Obstet Gynecol. Author manuscript; available in PMC 2013 November 01.

NIH

-PA

Author M

anuscriptN

IH-P

A A

uthor Manuscript

NIH

-PA

Author M

anuscript

NIH

-PA

Author M

anuscriptN

IH-P

A A

uthor Manuscript

NIH

-PA

Author M

anuscriptGrobman et al. Page 16

Table 5

Additional subgroup analyses for preterm birth prior to 37 and 34 weeks*

17-OHP(N=327)

Placebo(N=330)

RR (95% CI) P**

Preterm birth < 37 weeks

Cervical length 0.59

< 10 mm 5/9 (55.6) 10/16 (62.5) 0.89 (0.44 – 1.78)

10–20 mm 19/50 (38.0) 18/40 (45.0) 0.84 (0.52 – 1.38)

> 20 mm 58/268 (21.6) 52/274 (19.0) 1.14 (0.82 – 1.59)

Preterm birth < 34 weeks

Cervical length 0·82

< 15 mm 9/25 (36·0) 13/31(41.9) 0·86 (0·44–1·67)

≥15 mm 32/302 (10.6) 35/299 (11.7) 0.91 (0·58–1·42)

Cervical length 0·49

< 10 mm 5/9 (55·6) 6/16 (37·5) 1.48 (0·63–3.51)

10–20 mm 11/50 (22·0) 12/40 (30·0) 0·73 (0·36–1·48)

> 20 mm 25/268 (9.3) 30/274 (10.9) 0·85 (0·52–1·41)

*Data presented as n/N (%)

**P value for Breslow-Day interaction term

Am J Obstet Gynecol. Author manuscript; available in PMC 2013 November 01.

ALLGEMEINES

KRANKENHAUS




Romero et al., 2013

ALLGEMEINES

KRANKENHAUS




• Frauen mit zwei oder mehr Spätaborten oder Frühgeburten sind Kandidaten für eine ‘history-indicated’ Cerclage, welche typischerweise zwischen der SSW 12 und SSW14 gesetzt wird.

• Frauen mit einer Cervixinsuffizienz diagnostiziert durch eine manuelle Palpation zwischen der SSW 16 und 23 sind potentielle Kandidaten für eine ‘physical exam- indicated’ Cerclage.

• Frauen mit einer Einlingsschwangerschaft und einer vorrangegangene Frühgeburt und einer verkürzten Cervix (typischerweise <25mm) zwischen der SSW 16 und 23 sind Kandidaten für eine ‘ultrasound-indicated’ Cerclage.


Cervical stitch (cerclage) for preventing preterm birth insingleton pregnancy (Review)

Alfirevic Z, Stampalija T, Medley N

Alfirevic Z, Stampalija T, Medley N.

Cervical stitch (cerclage) for preventing preterm birth in singleton pregnancy.


DOI: 10.1002/14651858.CD008991.pub3.


Cervical stitch (cerclage) for preventing preterm birth in singleton pregnancy (Review)


Alfirevic et al., 2017

ALLGEMEINES

KRANKENHAUS





Cervical stitch (cerclage) for preventing preterm birth insingleton pregnancy (Review)

Alfirevic Z, Stampalija T, Medley N

Alfirevic Z, Stampalija T, Medley N.

Cervical stitch (cerclage) for preventing preterm birth in singleton pregnancy.


DOI: 10.1002/14651858.CD008991.pub3.


Cervical stitch (cerclage) for preventing preterm birth in singleton pregnancy (Review)


Analysis 1.8. Comparison 1 Cerclage versus no cerclage, Outcome 8 Preterm birth before 34 completed

weeks.

Review: Cervical stitch (cerclage) for preventing preterm birth in singleton pregnancy

Comparison: 1 Cerclage versus no cerclage

Outcome: 8 Preterm birth before 34 completed weeks

Study or subgroup Cerclage No cerclage Risk Ratio Weight Risk Ratio

n/N n/N

M-H,Random,95%

CI

M-H,Random,95%

CI

1 History-indicated cerclage vs no cerclage

Ezechi 2004 0/39 11/42 0.3 % 0.05 [ 0.00, 0.77 ]

MRC/RCOG 1993 92/635 113/629 36.0 % 0.81 [ 0.63, 1.04 ]

Rush 1984 14/96 14/98 4.9 % 1.02 [ 0.51, 2.03 ]

Subtotal (95% CI) 770 769 41.1 % 0.76 [ 0.40, 1.46 ]Total events: 106 (Cerclage), 138 (No cerclage)

Heterogeneity: Tau2 = 0.17; Chi2 = 4.66, df = 2 (P = 0.10); I2 =57%

Test for overall effect: Z = 0.81 (P = 0.42)

2 One-off ultrasound-indicated cerclage in high risk for PTL vs no cerclage

To 2004 6/26 11/30 3.2 % 0.63 [ 0.27, 1.46 ]


Heterogeneity: not applicable


3 Serial ultrasound-indicated cerclage in high risk for PTL vs no cerclage

Althuisius 2001 0/19 7/16 0.3 % 0.06 [ 0.00, 0.92 ]

Berghella 2004 10/25 11/22 5.6 % 0.80 [ 0.42, 1.51 ]

Owen 2009 42/148 57/153 21.2 % 0.76 [ 0.55, 1.06 ]

Rust 2000 13/61 15/66 5.3 % 0.94 [ 0.49, 1.81 ]


Heterogeneity: Tau2 = 0.03; Chi2 = 3.92, df = 3 (P = 0.27); I2 =23%


4 Physical exam-indicated cerclage in high risk for PTL vs no cerclage

Althuisius 2003 7/13 10/10 9.0 % 0.56 [ 0.34, 0.93 ]


Heterogeneity: not applicable


5 One-off ultrasound-indicated cerclage in low/unspecified risk for PTL vs no cerclage

Berghella 2004 0/3 1/7 0.3 % 0.67 [ 0.03, 12.96 ]

0.01 0.1 1 10 100

Favours cerclage Favours no cerclage

(Continued . . . )

82Cervical stitch (cerclage) for preventing preterm birth in singleton pregnancy (Review)


signifikant nur“physical exam- indicated”cerclage

ALLGEMEINES

KRANKENHAUS




Ehsanipoor et al.,2015

Review

Physical Examination–Indicated CerclageA Systematic Review and Meta-analysis

Robert M. Ehsanipoor, MD, Neil S. Seligman, MD, Gabriele Saccone, MD, Linda M. Szymanski, MD, PhD,Christina Wissinger, MS, MLIS, Erika F. Werner, MD, MS, and Vincenzo Berghella, MD

OBJECTIVE: To estimate the effectiveness of physicalexamination–indicated cerclage in the setting of second-trimester cervical dilatation by systematic review andmeta-analysis of published studies.

DATA SOURCES: We searched MEDLINE, EMBASE, Sco-pus, ClinicalTrials.gov, Web of Science, and the CochraneLibrary for studies published between 1966 and 2014 thatevaluated cervical cerclage for the treatment of cervicalinsufficiency.

METHODS OF STUDY SELECTION: The search yielded6,314 citations. We included cohort studies and random-ized controlled trials comparing cerclage placement withexpectant management of women with cervical dilata-tion between 14 and 27 weeks of gestation. Twoinvestigators independently reviewed each citation forinclusion or exclusion and discordant decisions werearbitrated by a third reviewer. Summary estimates werereported as the mean difference and 95% confidenceinterval (CI) for continuous variables or relative risk andwith 95% CI for dichotomous outcomes. Fixed- andrandom-effects meta-analysis was used, depending onheterogeneity.

TABULATION, INTEGRATION, AND RESULTS: Tenstudies met inclusion criteria and were included in thefinal analysis. One was a randomized controlled trial, twowere prospective cohort studies, and the remaining

seven were retrospective cohort studies. Of the 757women, 485 (64%) underwent physical examination–indicated cerclage placement and 272 (36%) were expec-tantly managed. Cerclage was associated with increasedneonatal survival (71% compared with 43%; relative risk1.65, 95% CI 1.19–2.28) and prolongation of pregnancy(mean difference 33.98 days, 95% CI 17.88–50.08).

CONCLUSION: Physical examination–indicated cerc-lage is associated with a significant increase in neonatalsurvival and prolongation of pregnancy of approximately1 month when compared with no such cerclage. Thestrength of this conclusion is limited by the potentialfor bias in the included studies.(Obstet Gynecol 2015;126:125–35)

DOI: 10.1097/AOG.0000000000000850

Cervical insufficiency, previously referred to ascervical incompetence, has classically been

defined as painless dilation of the cervix in the absenceof contractions or bleeding in the second trimester.1

Painless second-trimester cervical dilation is anuncommon finding in the general population occur-ring in less than 1% of pregnancies.2 Cerclage for theprevention of pregnancy loss in women with priorsecond-trimester loss or second-trimester cervical dila-tation in the index pregnancy was first reported in the1950s.3,4 Cerclage placement in the setting of cervicaldilatation has been variably referred to as “physicalexamination–indicated cerclage,” “rescue cerclage,”and “emergency cerclage.” To date, the benefits ofcerclage for this indication are not entirely clear.

The optimal evaluation and management ofasymptomatic patients presenting with second-trimester cervical dilatation remain controversial.There is only one randomized controlled trial to dateevaluating the use of cerclage in this clinical scenario,and it included only 23 patients, seven of whom werepregnant with twins.5 Several nonrandomized studieshave compared outcomes of women receiving cerc-lage with those expectantly managed in the setting

From the Departments of Gynecology and Obstetrics, Johns HopkinsUniversity, Baltimore, Maryland, University of Rochester, Rochester, NewYork, and Warren Alpert Medical School of Brown University, Providence,Rhode Island; the Department of Neuroscience, Reproductive Sciences andDentistry, School of Medicine, University of Naples Federico II, Naples, Italy;and the Division of Maternal-Fetal Medicine, Department of Obstetrics andGynecology, Sidney Kimmel Medical College of Thomas Jefferson University,Philadelphia, Pennsylvania.

Corresponding author: Robert M. Ehsanipoor, MD, Johns Hopkins UniversitySchool of Medicine, 600 N. Wolfe Street, Phipps 228, Baltimore, MD 21287;e-mail: [email protected].

Financial DisclosureThe authors did not report any potential conflicts of interest.

© 2015 by The American College of Obstetricians and Gynecologists. Publishedby Wolters Kluwer Health, Inc. All rights reserved.ISSN: 0029-7844/15

VOL. 126, NO. 1, JULY 2015 OBSTETRICS & GYNECOLOGY 125

Copyright ª by The American College of Obstetriciansand Gynecologists. Published by Wolters Kluwer Health, Inc.

Unauthorized reproduction of this article is prohibited.

In women undergoing cerclage, the incidence ofintraoperative membrane rupture was 4.1% (10 of246)5,12,19,21–23 and for cervical laceration was 7.9%(seven of 140)12,16,17,19; however, these data were notreported for the control groups. The patients in whichintraoperative membrane rupture occurred wereincluded in the treatment group based on the princi-ple of intention to treat for these 10 patients. Olatun-bosun also reported one case of intraoperativemembrane rupture, but this participant was excludedfrom the study.16 No maternal deaths were reported;however, one study reported an intensive care unitadmission for sepsis in a cerclage recipient,20 but thisoutcome was only specified in one other study.23

Rates of placental abruption, premature preterm rup-ture of membranes, and chorioamnionitis were incon-sistently and variably reported.

Six of the 10 studies were excluded from thesubanalysis of studies with the lowest risk of biasbecause a high risk of bias was noted in at least onecategory evaluated (Table 3).12,17,18,20,22,23 Althoughthe potential for treatment bias in the trial by Althui-sius et al was high, the decision was made to include

it in the subanalysis because indomethacin in thisclinical setting has not been shown to have a signifi-cant effect on outcomes.24 Of the remaining studies,outcome reporting by Ventolini et al21 was insuffi-cient to permit evaluation of bias for six of the sevencategories evaluated so it was excluded. Therefore,the studies by Daskalakis et al, Olatunbosun et al,and Althuisius et al were included in the subanaly-sis.5,16,19 In the subanalysis, cerclage was also associ-ated with higher rates of neonatal survival (78%compared with 33%, RR 2.11, 95% CI 1.41–3.55)and prolongation of pregnancy (mean difference34.00 days, 95% CI 3.11–64.89).

DISCUSSIONIncluded articles in this meta-analysis were limited innumber and variable in quality and study design.Although differences in our primary outcomes werefound, this meta-analysis also underscores the paucityand low quality of existing studies of physical exam-ination–indicated cerclage. Our findings suggest thatphysical examination–indicated cerclage is associatedwith significantly but modestly higher rates of neonatal

Table 5. Primary and Secondary Outcomes

Studied Outcome Studies/Participants Cerclage Expectant Effect Estimate*

Neonatal survival 8/657 294/413 (71) 106/244 (43) 1.66 (1.19–2.30)Delivery at less than 24 wk of gestation 3/316 51/221 (23) 31/95 (33) 0.47 (0.14–1.53)Delivery at 24—28 wk of gestation† 2/239 14/165 (8) 31/83 (37) 0.23 (0.13–0.41)Delivery at less than 34 wk of gestation 3/316 110/221 (50) 77/94 (82) 0.55 (0.38–0.80)Time to delivery (d) 6/385 56.7 18.8 33.98 (17.88–50.08)Gestational age at delivery (wk) 8/643 30.6 25.2 4.62 (3.89–5.36)Birth weight (g) 5/331 1,714.6 829.4 1,028 (714–1,341)

Data are n/N (%) or weighted mean unless otherwise specified.Bold indicates statistical significance.* Data are relative risk (95% confidence interval) for the top four rows and mean difference (95% confidence interval) for the bottom three

rows.† Referent group is women who delivered either before 24 weeks of gestation or after 28 weeks of gestation.

Fig. 1. Forest plot for neonatal survival. M-H, Mantel-Haenszel test; CI, confidence interval.Ehsanipoor. Physical Examination–Indicated Cerclage. Obstet Gynecol 2015.

132 Ehsanipoor et al Physical Examination–Indicated Cerclage OBSTETRICS & GYNECOLOGY

Copyright ª by The American College of Obstetriciansand Gynecologists. Published by Wolters Kluwer Health, Inc.

Unauthorized reproduction of this article is prohibited.

• 10 Studien eingeschlossen, 757 Patientinnen• 1 randomisierte Studie• 2 prospektive Studien

ALLGEMEINES

KRANKENHAUS




To et al., 2004

For personal use. Only reproduce with permission from The Lancet Publishing Group.

ARTICLES

THE LANCET • Vol 363 • June 5, 2004 • www.thelancet.com 1849

Summary

Background Cervical cerclage has been widely used in thepast 50 years to prevent early preterm birth and itsassociated neonatal mortality and morbidity. Results ofrandomised trials have not generally lent support to thispractice, but this absence of benefit may be due tosuboptimum patient selection, which was essentiallybased on obstetric history. A more effective way ofidentifying the high-risk group for early preterm deliverymight be by transvaginal sonographic measurement ofcervical length. We undertook a multicentre randomisedcontrolled trial to investigate whether, in women with ashort cervix identified by routine transvaginal scanning at22–24 weeks’ gestation, the insertion of a Shirodkarsuture reduces early preterm delivery.

Methods Cervical length was measured in 47 123 women.The cervix was 15 mm or less in 470, and 253 (54%) ofthese women participated in the study and wererandomised to cervical cerclage (127) or to expectantmanagement (126). Primary outcome was the frequency ofdelivery before 33 completed weeks (231 days) ofpregnancy.

Findings The proportion of preterm delivery before33 weeks was similar in both groups, 22% (28 of 127) inthe cerclage group versus 26% (33 of 126) in the controlgroup (relative risk=0·84, 95% CI 0·54–1·31, p=0·44), withno significant differences in perinatal or maternal morbidityor mortality.

Interpretation The insertion of a Shirodkar suture inwomen with a short cervix does not substantially reducethe risk of early preterm delivery. Routine sonographicmeasurement of cervical length at 22–24 weeks identifiesa group at high risk of early preterm birth.

Lancet 2004; 363: 1849–53

IntroductionPrematurity is the single largest factor in neonatalmortality, and is responsible for half of all neonataldeaths.1 Mortality rises from about 2% for infants bornat 32 weeks to more than 90% for those born at23 weeks.2 Moreover, handicap or disability arises inabout 60% of survivors after birth at 26 weeks and 30%in those born at 31 weeks.3 Furthermore, preterm birthis associated with a huge cost to the health servicebecause of both the need for intensive neonatal care,often for several weeks, and the continuing supportnecessary after discharge from hospital.4

Cervical cerclage has been used widely in themanagement of pregnancies regarded as at high risk ofpreterm delivery. Several observational studies in thepast 50 years have claimed high rates of successfulpregnancy outcome in women who had poor obstetrichistory attributed to cervical incompetence.5 However,the results of randomised trials have not consistentlysupported this practice, possibly because of suboptimumpatient selection, which was essentially based onobstetric history.6–8 A more effective way of identifying ahigh-risk group might be by transvaginal sonographicmeasurement of cervical length.9–12 A screening study at22–24 weeks has shown that the risk for spontaneousearly preterm delivery increases with decreasing cervicallength, from about 0·2% at 60 mm to 1·1% at 25 mm,4·0% at 15 mm, and 78% at 5 mm.10 Furthermore,demographic characteristics and obstetric history do nothave a substantial additional contribution to that ofcervical length in the prediction of preterm delivery.13

The aim of our multicentre randomised controlledtrial was to test the hypothesis that cervical cerclage inwomen found to have a short cervix on routinetransvaginal scanning at 22–24 weeks of gestation isassociated with a three-fold reduction in early pretermdelivery.

MethodsParticipants included 47 123 women with singletonpregnancies undergoing routine antenatal care in12 hospitals in the UK, Brazil, South Africa, Slovenia,Greece, and Chile. In these hospitals, women areroutinely offered an ultrasound scan between 11 weeksand 13 weeks 6 days for pregnancy dating and earlydiagnosis of major fetal defects and chromosomalabnormalities, and another scan between 22 weeks and24 weeks 6 days for examination of fetal anatomy andgrowth. As part of this study, women attending for the22–24 week scan were offered a transvaginal scan tomeasure cervical length, as a screening test forspontaneous preterm delivery.10 Women with a cervicallength of 15 mm or less were invited to participate in therandomised study of cervical cerclage. This cutoff wasselected because the risk of preterm delivery increasesexponentially below this length. All women gave

Cervical cerclage for prevention of preterm delivery in womenwith short cervix: randomised controlled trial

Meekai S To, Zarko Alfirevic, Victoria C F Heath, Simona Cicero, Ana Maria Cacho, Paula R Williamson, Kypros H Nicolaides,on behalf of the Fetal Medicine Foundation Second Trimester Screening Group*

*Members listed at end of report

Harris Birthright Research Centre for Fetal Medicine, King’sCollege Hospital Medical School, London, UK (M S To MRCOG,V C F Heath MB, S Cicero MD, A M Cacho MD,Prof K H Nicolaides MRCOG); and University Department of Obstetricsand Gynaecology, Liverpool Women’s Hospital(Prof Z Alfirevic MRCOG), and Centre for Medical Statistics and HealthEvaluation, University of Liverpool, Liverpool, UK(P R Williamson PhD)

Correspondence to: Prof K H Nicolaides, Harris Birthright ResearchCentre for Fetal Medicine, King’s College Hospital Medical School,Denmark Hill, London SE5 8RX (e-mail: [email protected])


ResultsFrom January, 1998, to May, 2002, 47 123 womenunderwent a transvaginal scan at 22–24 weeks’gestation. 470 had a short cervix and 253 (54%) of themagreed to participate in the study (figure). Table 1shows the baseline characteristics of the treatmentgroups. Women who declined participation did notdiffer from the study group in their main demographiccharacteristics and preterm delivery rate (data notshown).

Cervical cerclage was not associated with a significantreduction in the rate of early preterm delivery orneonatal mortality or morbidity compared withexpectant management (table 2). No heterogeneity oftreatment effects was evident across the participatingcentres. Delivery before 33 weeks was spontaneous in55 (90%) and iatrogenic in six (10%) cases. The relativeproportion of spontaneous delivery or delivery afterpreterm prelabour rupture of membranes in the twogroups was similar (30 of 33 [91%] in expectantlymanaged group vs 25 of 28 [89%] in cervical cerclagegroup). The indications for iatrogenic delivery before

33 weeks were fetal growth restriction, pre-eclampsia, orboth (one in the expectant group, two in the cerclagegroup), intrauterine death (one in the cerclage group),and abnormal antenatal fetal heart rate pattern (two inexpectant group). In the subgroup of women with ahistory of delivery at 16–32 weeks, six of 23 in thecerclage group delivered before 33 weeks, as did six of23 in the control group.

In the cerclage group, there were three stillbirths(born at 24–27 weeks), four neonatal deaths (born at26 weeks), and five babies with one or more neonatalmorbidities, all of whom were born before 27 weeks. Inthe expectantly managed group, there were fivestillbirths (four as a result of spontaneous delivery at23–24 weeks and one as a result of placental abruptionat 35 weeks), five neonatal deaths (born at 23–26 weeks)and seven babies with one or more neonatal morbidities,all of whom were born before 28 weeks.

For the cerclage group, mean birthweight was 204 ggreater than that of the controls. However, ANCOVAshowed that this difference was merely a consequence ofthe 1-week difference in gestational age at delivery (datanot shown). The use of antibiotics was not aprespecified outcome; we noted that 35 of 127 (28%)women in the cerclage group received antibiotics forsymptomatic vaginal discharge, prelabour rupture ofmembranes, suspected chorioamnionitis, or longprophylaxis compared with 15 of 126 (12%) in thecontrol group. The main indications for antibiotics wereprolonged prophylaxis after cerclage and symptomaticvaginal discharge, which was significantly greater inthose with a suture than in those managed expectantly(table 2). In the cerclage group, tocolytics were usedimmediately after surgery or because of threatenedpreterm labour in eight of 127 (6%) of womencompared with four of 126 (3%) in the control group.Emergency removal of the cervical cerclage wasnecessary in 41 of 122 women (34%), including 24 casesbefore 33 weeks because of preterm labour, in ten casesat 33–37 weeks because of preterm labour or rupture ofmembranes, and in seven cases after 37 weeks whenlabour began.

Seven women did not receive the treatment they wereallocated. In the control group, two underwent cerclageon maternal request, one at 25 weeks and the other at26 weeks gestation. They both delivered spontaneouslyat 30 weeks 4 days and 36 weeks 5 days, respectively. In

ARTICLES


Cerclage (n=127) Expectant (n=126)

Maternal age (years) 29·8 (14·7–43·0) 29·3 (13·9–41·2)Gestation at entry (weeks) 23·5 (22·3–25·9) 23·6 (22·3–25·3)Body-mass index (kg/m2) 26·5 (16·6–42·2) 26·0 (15·9–46·1)Smoker 10 (8%) 17 (13%)Obstetric history

Primigravida 32 (25%) 33 (26%)All previous deliveries 27 (21%) 16 (13%)after 32 weeksAll pregnancy losses 45 (35%) 54 (43%)before 16 weeksOne delivery at 21 (17%) 16 (13%)16–32 weeksTwo or more deliveries 2 (2%) 7 (6%)at 16–32 weeks

Ethnic originBlack 62 (49%) 73 (58%)White 59 (46%) 48 (38%)Asian 6 (5%) 5 (4%)

Previous cervical surgeryKnife cone or loop excision 5 (4%) 7 (6%)Previous cerclage 2 (2%) 2 (2%)

Cervical length (mm) 9·6 (2–15) 9·3 (2–15)Cervical funnelling (yes) 121 (95%) 117 (93%)

Data are mean (range) or number (%) unless otherwise indicated.

Table 1: Baseline characteristics

Cerclage (n=127) Expectant (n=126) RR (95% CI) p

Delivery before 33 weeks 28 (22%) 33 (26%) 0·84 (0·54 to 1·31) 0·4411Gestation at delivery (weeks) (mean [SE]) 36·4 (0·42) 35·4 (0·45) 0·95 (–0·26 to 2·15)* 0·1233GA adjusted BW centile† (median) 41·8 (1·1–99·8) 37·2 (0·7–100) 0·0983Birthweight‡ (g) (mean [SE]) 2769 (85) 2565 (85) 204 (–33 to 441)* 0·0917Onset of labour

Spontaneous 92 (72%) 104 (83%) 0·88 (0·77 to 1·00) 0·0692Iatrogenic 28 (22%) 18 (14%) 1·54 (0·90 to 2·64)Indicated after PPROM 7 (6%) 4 (3%) 1·74 (052 to 5·78)

Caesarean section 33 (26%) 23 (18%) 1·42 (0·89 to 2·28) 0·1386Preterm membrane rupture 23 (18%) 19 (15%) 1·20 (0·69 to 2·09) 0·5171Maternal pyrexia‡ 5 (4%) 1 (1%) 4·92 (0·58 to 41·53) 0·2132Symptomatic vaginal discharge‡ 8 (6%) 1 (1%) 7·87 (1·00 to 62·04) 0·0358Stillbirth 3 (2%) 5 (4%) 0·60 (0·15 to 2·44) 0·4999Perinatal deaths§ 7 (6%) 10 (8%) 0·69 (0·27 to 1·77) 0·4412Positive fetal blood culture¶ 5/123 (4%) 2/121 (2%) 2·46 (0·49 to 12·43) 0·4464Bronchopulmonary dysplasia¶ 4/123 (3%) 4/121 (3%) 0·98 (0·25 to 3·84) 1·0000IVH/PVH¶ 1/123 (1%) 2/121 (2%) 0·49 (0·05 to 5·35) 0·6204Retinopathy of prematurity¶ 0/123 (0%) 3/121 (2%) 0·14 (0·01 to 2·69) 0·1204

GA=gestational age. BW=birthweight. PPROM=preterm prelabour rupture of membranes. IVH=intraventricular haemorrhage. PVH=periventricular haemorrhage. Data arenumber (%) unless otherwise indicated. *Difference in means (95% CI). †Data missing for one infant in cerclage group and for three infants in expectant group.‡Outcome data missing for one woman in expectant group. §Total number of stillbirths and neonatal deaths before hospital discharge. ¶Stillbirths excluded fromcalculations: missing data for one neonate in cerclage group. Tested where clinically indicated, otherwise assumed to be negative.

Table 2: Primary analyses of outcome measures


ResultsFrom January, 1998, to May, 2002, 47 123 womenunderwent a transvaginal scan at 22–24 weeks’gestation. 470 had a short cervix and 253 (54%) of themagreed to participate in the study (figure). Table 1shows the baseline characteristics of the treatmentgroups. Women who declined participation did notdiffer from the study group in their main demographiccharacteristics and preterm delivery rate (data notshown).

Cervical cerclage was not associated with a significantreduction in the rate of early preterm delivery orneonatal mortality or morbidity compared withexpectant management (table 2). No heterogeneity oftreatment effects was evident across the participatingcentres. Delivery before 33 weeks was spontaneous in55 (90%) and iatrogenic in six (10%) cases. The relativeproportion of spontaneous delivery or delivery afterpreterm prelabour rupture of membranes in the twogroups was similar (30 of 33 [91%] in expectantlymanaged group vs 25 of 28 [89%] in cervical cerclagegroup). The indications for iatrogenic delivery before

33 weeks were fetal growth restriction, pre-eclampsia, orboth (one in the expectant group, two in the cerclagegroup), intrauterine death (one in the cerclage group),and abnormal antenatal fetal heart rate pattern (two inexpectant group). In the subgroup of women with ahistory of delivery at 16–32 weeks, six of 23 in thecerclage group delivered before 33 weeks, as did six of23 in the control group.

In the cerclage group, there were three stillbirths(born at 24–27 weeks), four neonatal deaths (born at26 weeks), and five babies with one or more neonatalmorbidities, all of whom were born before 27 weeks. Inthe expectantly managed group, there were fivestillbirths (four as a result of spontaneous delivery at23–24 weeks and one as a result of placental abruptionat 35 weeks), five neonatal deaths (born at 23–26 weeks)and seven babies with one or more neonatal morbidities,all of whom were born before 28 weeks.

For the cerclage group, mean birthweight was 204 ggreater than that of the controls. However, ANCOVAshowed that this difference was merely a consequence ofthe 1-week difference in gestational age at delivery (datanot shown). The use of antibiotics was not aprespecified outcome; we noted that 35 of 127 (28%)women in the cerclage group received antibiotics forsymptomatic vaginal discharge, prelabour rupture ofmembranes, suspected chorioamnionitis, or longprophylaxis compared with 15 of 126 (12%) in thecontrol group. The main indications for antibiotics wereprolonged prophylaxis after cerclage and symptomaticvaginal discharge, which was significantly greater inthose with a suture than in those managed expectantly(table 2). In the cerclage group, tocolytics were usedimmediately after surgery or because of threatenedpreterm labour in eight of 127 (6%) of womencompared with four of 126 (3%) in the control group.Emergency removal of the cervical cerclage wasnecessary in 41 of 122 women (34%), including 24 casesbefore 33 weeks because of preterm labour, in ten casesat 33–37 weeks because of preterm labour or rupture ofmembranes, and in seven cases after 37 weeks whenlabour began.

Seven women did not receive the treatment they wereallocated. In the control group, two underwent cerclageon maternal request, one at 25 weeks and the other at26 weeks gestation. They both delivered spontaneouslyat 30 weeks 4 days and 36 weeks 5 days, respectively. In

ARTICLES


Cerclage (n=127) Expectant (n=126)

Maternal age (years) 29·8 (14·7–43·0) 29·3 (13·9–41·2)Gestation at entry (weeks) 23·5 (22·3–25·9) 23·6 (22·3–25·3)Body-mass index (kg/m2) 26·5 (16·6–42·2) 26·0 (15·9–46·1)Smoker 10 (8%) 17 (13%)Obstetric history

Primigravida 32 (25%) 33 (26%)All previous deliveries 27 (21%) 16 (13%)after 32 weeksAll pregnancy losses 45 (35%) 54 (43%)before 16 weeksOne delivery at 21 (17%) 16 (13%)16–32 weeksTwo or more deliveries 2 (2%) 7 (6%)at 16–32 weeks

Ethnic originBlack 62 (49%) 73 (58%)White 59 (46%) 48 (38%)Asian 6 (5%) 5 (4%)

Previous cervical surgeryKnife cone or loop excision 5 (4%) 7 (6%)Previous cerclage 2 (2%) 2 (2%)

Cervical length (mm) 9·6 (2–15) 9·3 (2–15)Cervical funnelling (yes) 121 (95%) 117 (93%)

Data are mean (range) or number (%) unless otherwise indicated.

Table 1: Baseline characteristics

Cerclage (n=127) Expectant (n=126) RR (95% CI) p

Delivery before 33 weeks 28 (22%) 33 (26%) 0·84 (0·54 to 1·31) 0·4411Gestation at delivery (weeks) (mean [SE]) 36·4 (0·42) 35·4 (0·45) 0·95 (–0·26 to 2·15)* 0·1233GA adjusted BW centile† (median) 41·8 (1·1–99·8) 37·2 (0·7–100) 0·0983Birthweight‡ (g) (mean [SE]) 2769 (85) 2565 (85) 204 (–33 to 441)* 0·0917Onset of labour

Spontaneous 92 (72%) 104 (83%) 0·88 (0·77 to 1·00) 0·0692Iatrogenic 28 (22%) 18 (14%) 1·54 (0·90 to 2·64)Indicated after PPROM 7 (6%) 4 (3%) 1·74 (052 to 5·78)

Caesarean section 33 (26%) 23 (18%) 1·42 (0·89 to 2·28) 0·1386Preterm membrane rupture 23 (18%) 19 (15%) 1·20 (0·69 to 2·09) 0·5171Maternal pyrexia‡ 5 (4%) 1 (1%) 4·92 (0·58 to 41·53) 0·2132Symptomatic vaginal discharge‡ 8 (6%) 1 (1%) 7·87 (1·00 to 62·04) 0·0358Stillbirth 3 (2%) 5 (4%) 0·60 (0·15 to 2·44) 0·4999Perinatal deaths§ 7 (6%) 10 (8%) 0·69 (0·27 to 1·77) 0·4412Positive fetal blood culture¶ 5/123 (4%) 2/121 (2%) 2·46 (0·49 to 12·43) 0·4464Bronchopulmonary dysplasia¶ 4/123 (3%) 4/121 (3%) 0·98 (0·25 to 3·84) 1·0000IVH/PVH¶ 1/123 (1%) 2/121 (2%) 0·49 (0·05 to 5·35) 0·6204Retinopathy of prematurity¶ 0/123 (0%) 3/121 (2%) 0·14 (0·01 to 2·69) 0·1204

GA=gestational age. BW=birthweight. PPROM=preterm prelabour rupture of membranes. IVH=intraventricular haemorrhage. PVH=periventricular haemorrhage. Data arenumber (%) unless otherwise indicated. *Difference in means (95% CI). †Data missing for one infant in cerclage group and for three infants in expectant group.‡Outcome data missing for one woman in expectant group. §Total number of stillbirths and neonatal deaths before hospital discharge. ¶Stillbirths excluded fromcalculations: missing data for one neonate in cerclage group. Tested where clinically indicated, otherwise assumed to be negative.

Table 2: Primary analyses of outcome measures

ALLGEMEINES

KRANKENHAUS




Berghella et al., 2010

Ultrasound Obstet Gynecol 2010; 35: 468–473Published online 5 January 2010 in Wiley InterScience (www.interscience.wiley.com). DOI: 10.1002/uog.7547

Effectiveness of cerclage according to severity of cervicallength shortening: a meta-analysis

V. BERGHELLA*, S. M. KEELER†, M. S. TO‡, S. M. ALTHUISIUS§ and O. A. RUST¶

Divisions of Maternal-Fetal Medicine, Departments of Obstetrics and Gynecology, *Jefferson Medical College of Thomas JeffersonUniversity, Philadelphia and ¶Lehigh Valley Hospital and Health Network, Allentown, PA and †New York University, New York, NY,USA, ‡St Thomas’s and Guy’s Hospital, London, UK and §Department of Obstetrics and Gynecology, Maasstad Ziekenhuis, Rotterdam,The Netherlands

KEYWORDS: cerclage; cervical length; preterm birth; transvaginal ultrasound

ABSTRACT

Objectives To estimate the effectiveness of cerclageaccording to degree of cervical length (CL) shortening.

Methods A meta-analysis was carried out of trials ofwomen with singleton gestations and second-trimestertransvaginal sonographic CL < 25 mm randomized tocerclage or no cerclage. The degree of CL shortening wascorrelated to the efficacy of cerclage in preventing pretermbirth.

Results There was a significant reduction in preterm birth< 35 weeks in the cerclage compared with no cerclagegroups in 208 singleton gestations with both a previ-ous preterm birth and CL < 25 mm (relative risk, 0.61;95% CI, 0.40–0.92). In these women, preterm birth< 37 weeks was significantly reduced with cerclage forCL ≤ 5.9 mm, ≤ 15.9 mm, 16–24.9 mm and < 25 mm.None of the analyses for 344 women without a previouspreterm birth was significant.

Conclusions Cerclage, when performed in women witha singleton gestation, previous preterm birth and cervicallength < 25 mm, seems to have a similar effect regardlessof the degree of cervical shortening, including CL16–24 mm, as well as CL ≤ 5.9 mm. Copyright © 2010ISUOG. Published by John Wiley & Sons, Ltd.

INTRODUCTION

Preterm birth is the main cause of perinatal morbidity andmortality. A short cervical length (CL) on transvaginalultrasound examination in the second trimester predictspreterm birth. The shorter the CL, the higher the risk

of delivering preterm. Ultrasound-indicated cerclage hasbeen proposed to prevent preterm birth once a shortcervix has been detected on ultrasound imaging. A meta-analysis1 of the first four trials2–5 and a recent NationalInstitute of Child Health and Human Development-sponsored trial6 have reported benefit of this interventionin women with a singleton gestation, previous pretermbirth and short CL.

The degree of cervical shortening has varied in these fivetrials between ≤ 15 mm and < 25 mm2–6. It is unknownwhether the efficacy of ultrasound-indicated cerclagevaries depending on the degree of cervical shortening,as some have postulated that cerclage may not be veryefficacious for the very short cervix (e.g. < 6 mm) ormildly shortened cervix (e.g. 15–24 mm)6.

Our aim was to estimate the effectiveness of cerclagefor a short cervix by degrees of CL shortening.

METHODS

Literature search

We searched MEDLINE, PubMed, EMBASE and theCochrane Library for randomized trials on cerclagefor short CL on transvaginal ultrasound examinationpublished between 1966 and June 2008. Search termswere: ‘cerclage’, ‘short cervix’, ‘ultrasound’, ‘randomizedtrial’. There were no restrictions based on language. Allmanuscripts were reviewed for pertinent references.

Selection

We included randomized trials involving the use of cer-clage in women with a short CL on second-trimester

Correspondence to: Dr V. Berghella, Division of Maternal–Fetal Medicine, Department of Obstetrics and Gynecology, Jefferson MedicalCollege of Thomas Jefferson University, 834 Chestnut Street, Suite 400, Philadelphia, PA 19107, USA(e-mail: [email protected])

Accepted: 9 July 2009

Copyright © 2010 ISUOG. Published by John Wiley & Sons, Ltd. ORIGINAL PAPER

470 Berghella et al.

Table 2 Incidence of preterm birth < 35 weeks according to degree of cervical length (CL) shortening and clinical characteristics

Preterm birth < 35 weeks (n (%))

Cerclage No cerclage RR (95% CI)

AllCL ≤ 5.9 mm 22/45 (48.9) 25/35 (71.4) 0.68 (0.50–0.99)†CL 6–10.9 mm 19/65 (29.2) 19/54 (35.2) 0.83 (0.50–1.40)CL 11–15.9 mm 14/74 (18.9) 27/95 (28.4) 0.67 (0.37–1.16)CL 16–20.9 mm 6/37 (16.2) 6/25 (24.0) 0.68 (0.25–1.83)CL 21–24.9 mm 8/57 (14.0) 15/65 (23.1) 0.61 (0.28–1.29)CL ≤ 15.9 mm 55/184 (29.9) 71/184 (38.6) 0.77 (0.58–1.03)CL 16–24.9 mm 14/94 (14.9) 21/90 (23.3) 0.64 (0.35–1.18)CL < 25 mm* 69/278 (24.8) 92/274 (33.6) 0.74 (0.57–0.96)†

No previous PTBCL ≤ 5.9 mm 17/35 (48.6) 17/26 (65.4) 0.74 (0.49–1.15)CL 6–10.9 mm 17/53 (32.1) 14/41 (34.1) 0.94 (0.53–1.69)CL 11–15.9 mm 7/54 (13.0) 18/72 (25.0) 0.49 (0.22–1.04)CL 16–20.9 mm 2/14 (14.3) 1/8 (12.5) 1.14 (0.16–8.47)CL 21–24.9 mm 1/15 (6.7) 3/26 (11.5) 0.58 (0.08–3.74)CL ≤ 15.9 mm 41/142 (28.9) 49/139 (35.3) 0.82 (0.58–1.15)CL 16–24.9 mm 3/29 (10.3) 4/34 (11.8) 0.88 (0.23–3.32)CL < 25 mm* 44/171 (25.7) 53/173 (30.6) 0.84 (0.60–1.18)

Previous PTB < 37 weeksCL ≤ 5.9 mm 5/10 (50.0) 8/9 (88.9) 0.56 (0.29–1.09)CL 6–10.9 mm 2/12 (16.7) 5/13 (38.5) 0.43 (0.11–1.59)CL 11–15.9 mm 7/20 (35.0) 9/23 (39.1) 0.89 (0.41–1.96)CL 16–20.9 mm 4/23 (17.4) 5/17 (29.4) 0.59 (0.19–1.81)CL 21–24.9 mm 7/42 (16.7) 12/39 (30.8) 0.54 (0.24–1.23)CL ≤ 15.9 mm 14/42 (33.3) 22/45 (48.9) 0.68 (0.40–1.15)CL 16–24.9 mm 11/65 (16.9) 17/56 (30.4) 0.56 (0.29–1.07)CL < 25 mm* 25/107 (23.4) 39/101 (38.6) 0.61 (0.40–0.92)†

*Includes CL ≥ 25 mm but also with ≥ 25% funneling. †Significant results. PTB, preterm birth; RR, relative risk.

Table 3 Incidence of preterm birth < 37 weeks according to degree of cervical length (CL) shortening and clinical characteristics

Preterm birth < 37 weeks (n (%))

Cerclage No cerclage RR (95% CI)

AllCL ≤ 5.9 mm 26/45 (57.8) 29/35 (82.9) 0.70 (0.52–0.93)†CL 6–10.9 mm 25/65 (38.5) 27/54 (50.0) 0.77 (0.51–1.16)CL 11–15.9 mm 24/74 (32.4) 40/95 (42.1) 0.77 (0.51–1.16)CL 16–20.9 mm 14/37 (37.8) 14/25 (56.0) 0.68 (0.39–1.16)CL 21–24.9 mm 19/57 (33.3) 25/65 (38.5) 0.87 (0.54–1.40)CL ≤ 15.9 mm 75/184 (40.8) 96/184 (52.2) 0.78 (0.63–0.98)†CL 16–24.9 mm 33/94 (35.1) 40/90 (44.4) 0.79 (0.55–1.13)CL < 25 mm* 108/278 (38.8) 135/274 (49.3) 0.79 (0.65–0.95)†

No previous PTBCL ≤ 5.9 mm 20/35 (57.1) 20/26 (76.9) 0.74 (0.52–1.06)CL 6–10.9 mm 20/53 (37.7) 19/41 (46.3) 0.81 (0.51–1.31)CL 16–20.9 mm 5/14 (35.7) 3/8 (37.5) 0.95 (0.31–2.97)CL 21–24.9 mm 6/15 (40.0) 6/26 (23.1) 1.73 (0.68–4.42)CL ≤ 10.9 mm 40/88 (45.5) 39/67 (58.2) 0.78 (0.58–1.06)CL ≤ 15.9 mm 56/142 (39.4) 65/139 (46.8) 0.84 (0.64–1.11)CL 10–20.9 mm 36/86 (41.9) 35/96 (36.5) 1.15 (0.80–1.65)CL 16–24.9 mm 11/29 (37.9) 10/34 (29.4) 1.29 (0.64–2.59)CL < 25 mm* 67/171 (39.2) 74/173 (42.8) 0.92 (0.71–1.18)

Previous PTB < 37 weeksCL ≤ 5.9 mm 6/10 (60.0) 9/9 (100) 0.60 (0.36–0.99)†CL 6–10.9 mm 5/12 (41.7) 8/13 (61.5) 0.73 (0.32–1.64)CL 11–15.9 mm 8/20 (40.0) 14/23 (60.9) 0.66 (0.35–1.23)CL 16–20.9 mm 9/23 (39.1) 11/17 (64.7) 0.60 (0.33–1.12)CL 21–24.9 mm 13/42 (30.9) 19/39 (48.7) 0.64 (0.36–1.11)CL ≤ 15.9 mm 19/42 (45.2) 31/45 (68.9) 0.66 (0.45–0.97)†CL 16–24.9 mm 22/65 (33.8) 30/56 (53.6) 0.63 (0.42–0.96)†CL < 25 mm* 41/107 (38.3) 61/101 (60.4) 0.63 (0.48–0.85)†

*Includes CL ≥ 25 mm but also with ≥ 25% funneling. †Significant results. PTB, preterm birth; RR, relative risk.

Copyright © 2010 ISUOG. Published by John Wiley & Sons, Ltd. Ultrasound Obstet Gynecol 2010; 35: 468–473.

Cerclage effectiveness by cervical length 473

0

10

20

30

40

50

60

70

80

90(a)

(b)

Pre

term

bir

th <

35 w

eeks

(%)

≤ 5.9 6−10.9 11−15.9 16−20.9 21−24.9

Cervical length (mm)

0

10

20

30

40

50

60

70

Pre

term

bir

th <

35 w

eeks

(%)

≤ 5.9 6−10.9 11−15.9 16−20.9 21−24.9

Cervical length (mm)

Figure 2 Incidence of preterm birth < 35 weeks as affected bycerclage (!) or no cerclage ( ), stratified by different cervical lengthcut-offs in women with singleton gestations, a short cervical lengthand a previous preterm birth (a) or no previous preterm birth (b).

0.56), < 32 weeks (RR, 0.53) and < 28 weeks (RR, 0.62),with the RR at < 24 weeks even lower at 0.22.

The issue of the effect of cerclage in women without aprevious preterm birth and a short CL is still unresolved.If there is an effect, it not going to be as strong as thatin women with a previous preterm birth. Although thereductions (many significant) in preterm birth for womenwith a previous preterm birth are often 30–40%, thereductions would be at most 10–30% in women with-out a previous preterm birth (Tables 2–6); no significantbenefit was seen in any of these analyses in this group oflower-risk women without a poor obstetric history.

The main shortcoming of our study is the lack of powerto estimate an effect in all these subgroups. Our studyshould be regarded as a pilot to guide future researchstudies. Despite this limitation, the trends, especially forwomen with a previous preterm birth, were remarkablysimilar. We did not look at other outcomes such as peri-natal mortality in this study because these have beenreported already in the primary analysis, and the num-bers were too small for subanalyses. Another limitationof the interpretation of our findings is the number ofstatistical comparisons that were done in the course ofexploring the data (Tables 2–6). As significance couldoccur by chance once in about 20 comparisons, CIs andsignificance values other than those for the primary analy-sis among all patients should be interpreted with caution.Our meta-analysis, using individual patient-level data, hasthe advantage of providing information about potential

predictors of outcomes such as degree of cervical shorten-ing which are not available using summary statistics frompublished data.

Given that women with singleton gestations with aprevious preterm birth and a CL < 25 mm seem to ben-efit from cerclage, waiting until a shorter CL developsbefore offering cerclage may not be necessary, if theseresults are confirmed. The number of women with a pre-vious preterm birth and a CL < 25 mm who should betreated with cerclage to prevent one case of preterm birth< 35 weeks is approximately eight. Based on the NIH trialand previous results, we suggest screening women witha previous spontaneous preterm birth with transvaginalultrasound CL measurements starting at 16 weeks, thenevery 2 weeks until 23 weeks.

Shirodkar8 and MacDonald9 both reported more than50 years ago that cerclage was indicated for womenwho had both of the following characteristics: previ-ous preterm birth (or second-trimester loss) and change incervical examination. We argue, now that cerclage seemsbeneficial in this population, that the diagnosis of cer-vical insufficiency should be ‘previous preterm birth (orsecond-trimester loss) in a woman currently carrying a sin-gleton gestation and with second-trimester (14–24 weeks)asymptomatic cervical shortening (< 25 mm) shown bytransvaginal ultrasound imaging or cervical dilatationconfirmed by physical examination’.

REFERENCES

1. Berghella V, Odibo AO, To MS, Rust OA, Althiusius SM. Cer-clage for short cervix on ultrasound: Meta-analysis of trialsusing individual patient-level data. Obstet Gynecol 2005; 106:181–189.

2. Rust OA, Atlas RO, Reed J, van Gaalen J, Balducci J. Revisitingthe short cervix detected by transvaginal ultrasound in the secondtrimester: why cerclage therapy may not help. Am J ObstetGynecol 2001; 185: 1098–1105.

3. Althuisius SM, Dekker GA, Hummel P, Bekedam DJ, vanGeijn HP. Final results of the cervical incompetence preventionrandomized cerclage trial (CIPRACT): therapeutic cerclage withbed rest versus bed rest alone. Am J Obstet Gynecol 2001; 185:1106–1112.

4. To MS, Alfirevic Z, Heath VCF, Cacho AM, Williamson PR,Nicolaides KH. Cervical cerclage for prevention of pretermdelivery in women with short cervix: randomized controlledtrial. Lancet 2004; 363: 1849–1853.

5. Berghella V, Odibo AO, Tolosa JE. Cerclage for prevention ofpreterm birth in women with a short cervix found on transvaginalultrasound examination: a randomized trial. Am J ObstetGynecol 2004; 191: 1311–1317.

6. Owen J, for the Vaginal Ultrasound Trial Consortium. Multicen-ter randomized trial of cerclage for preterm birth prevention inhigh-risk women with shortened mid-trimester cervical length.Am J Obstet Gynecol 2009; 201: 375.e1–375.e8.

7. Moher D, Cook DJ, Eastwood S, Oklin I, Rennie D, Stroup DF,for the QUOROM Group. Improving the quality of reports ofmeta-analyses of randomized controlled trials: the QUOROMstatement. Lancet 1999; 354: 1896–1900.

8. Shirodkar VN. A new method of operative treatment for habitualabortions in the second trimester of pregnancy. Antiseptic 1955;52: 299–300.

9. McDonald IA. Suture of the cervix for inevitable miscarriage.J Obstet Gynaecol Br Emp 1957; 64: 346–350.

Copyright © 2010 ISUOG. Published by John Wiley & Sons, Ltd. Ultrasound Obstet Gynecol 2010; 35: 468–473.

ALLGEMEINES

KRANKENHAUS




Conde-Agudelo et al., 2018

Vaginal progesterone is as effective as cervical cerclageto prevent preterm birth in women with a singletongestation, previous spontaneous preterm birth,and a short cervix: updated indirect comparisonmeta-analysisAgustin Conde-Agudelo, MD, MPH, PhD; Roberto Romero, MD, DMedSci; Eduardo Da Fonseca, MD; John M. O’Brien, MD;Elcin Cetingoz, MD; George W. Creasy, MD; Sonia S. Hassan, MD; Offer Erez, MD; Percy Pacora, MD;Kypros H. Nicolaides, MD

BACKGROUND: An indirect comparison meta-analysis published in2013 reported that both vaginal progesterone and cerclage are equallyefficacious for preventing preterm birth and adverse perinatal outcomes inwomen with a singleton gestation, previous spontaneous preterm birth,and a sonographic short cervix. The efficacy of vaginal progesterone hasbeen challenged after publication of the OPPTIMUM study. However, thishas been resolved by an individual patient-data meta-analysis (Am JObstet Gynecol. 2018;218:161-180).OBJECTIVE: To compare the efficacy of vaginal progesterone andcerclage in preventing preterm birth and adverse perinatal outcomes inwomen with a singleton gestation, previous spontaneous preterm birth,and a midtrimester sonographic short cervix.DATASOURCES:MEDLINE, EMBASE, LILACS, and CINAHL (from theirinception to March 2018); Cochrane databases, bibliographies, andconference proceedings.STUDY ELIGIBILITY CRITERIA: Randomized controlled trialscomparing vaginal progesterone to placebo/no treatment or cerclage to nocerclage in women with a singleton gestation, previous spontaneouspreterm birth, and a sonographic cervical length <25 mm.STUDY APPRAISAL AND SYNTHESIS METHODS: Updatedsystematic review and adjusted indirect comparison meta-analysis ofvaginal progesterone vs cerclage using placebo/no cerclage as thecommon comparator. The primary outcomes were preterm birth <35weeks of gestation and perinatal mortality. Pooled relative risks (RRs) with95% confidence intervals were calculated.

RESULTS: Five trials comparing vaginal progesterone vs placebo (265women) and 5 comparing cerclage vs no cerclage (504 women) wereincluded. Vaginal progesterone, compared to placebo, significantlyreduced the risk of preterm birth <35 and <32 weeks of gestation,composite perinatal morbidity/mortality, neonatal sepsis, compositeneonatal morbidity, and admission to the neonatal intensive care unit (RRsfrom 0.29 to 0.68). Cerclage, compared to no cerclage, significantlydecreased the risk of preterm birth <37, <35, <32, and <28 weeks ofgestation, composite perinatal morbidity/mortality, and birthweight<1500 g (RRs from 0.64 to 0.70). Adjusted indirect comparison meta-analyses did not show statistically significant differences betweenvaginal progesterone and cerclage in the reduction of preterm birth oradverse perinatal outcomes.CONCLUSION: Vaginal progesterone and cerclage are equallyeffective for preventing preterm birth and improving perinatal out-comes in women with a singleton gestation, previousspontaneous preterm birth, and a midtrimester sonographic shortcervix. The choice of treatment will depend on adverse events andcost-effectiveness of interventions and patient/physician’spreferences.

Key words: admission to neonatal intensive care unit, birthweight<1500 g, cervical length, cervical stitch, perinatal mortality, prematurity,progestin, progestogens, recurrent preterm birth, transvaginal ultrasound,uterine cervix

W orldwide, an estimated 11.1% ofall live births in 2010 were deliv-

ered preterm (14.9 million babies).1 Inthe United States, the preterm birth ratehad declined steadily from 2007 to 2014.In 2016, the rate of preterm birth rose to9.85%, a 2% rise from 2015 and the sec-ond straight year of increase for this rate.2

Complications of preterm birth arethe leading cause of neonatal mortality,

responsible for 35% of the world’s 2.6million deaths that occurred in 2016.3 Inaddition to its contribution to neonataland child morbidity and mortality, pre-term birth has lifelong effects on neu-rodevelopmental functioning such asincreased risk of cerebral palsy, intellec-tual disability, visual and hearing im-pairments, and an increased risk ofchronic disease in adulthood.4-11

It is widely accepted that preterm birthis a syndrome caused by several patho-logical processes such as infection,vascular and decidual disorders, uterine

overdistension, breakdown of maternal-fetal tolerance, a decline in progesteroneaction, and cervical disease.12-14 A previ-ous spontaneous preterm birth is awell-known risk factor for recurrentspontaneous preterm delivery.15-27 Arecent meta-analysis reported that theoverall risk of recurrent spontaneouspreterm birth<37 weeks of gestation was30%.28 A short cervix, conventionallydefined as a transvaginal sonographiccervical length !25mm in the mid-trimester of pregnancy, is also an impor-tant risk factor for spontaneous pretermdelivery and has emerged as one of thestrongest and most consistent predictors

0002-9378/freePublished by Elsevier Inc.https://doi.org/10.1016/j.ajog.2018.03.028

Related editorial, page 5.

10 American Journal of Obstetrics & Gynecology JULY 2018

Report of Major Impact ajog.org

TABLE 2Direct comparisons: vaginal progesterone vs placebo and cerclage vs no cerclage

Outcome

Vaginal progesterone vs placebo Cerclage vs no cerclage

Trials,n

Vaginalprogesterone Placebo RR (95% CI)

Heterogeneity_2, %

NNT(95% CI) Trials, n Cerclage No cerclage RR (95% CI)

HeterogeneityP valuea

NNT(95% CI)

Primary outcomes

Preterm birth<35 weeks

5 44/139 (32%) 58/126 (46%) 0.68(0.50e0.93)

0 7 (4e31) 5 71/250 (28%) 105/254 (41%) 0.70 (0.55e0.89) !0.10 8 (5e22)

Perinatalmortality

5 7/139 (5%) 10/126 (8%) 0.63(0.26e1.56)

0 — 5 22/250 (9%) 35/254 (14%) 0.65 (0.40e1.07) !0.10 —

Secondaryoutcomes


5 67/139 (48%) 74/126 (59%) 0.82(0.65e1.02)

0 — 5 105/250 (42%) 154/254 (61%) 0.70 (0.58e0.83) !0.10 6 (4e10)


5 27/139 (19%) 40/126 (32%) 0.60(0.39e0.92)

0 8 (5e39) 5 48/250 (19%) 75/254 (30%) 0.66 (0.48e0.91) !0.10 10 (7e38)


5 18/139 (13%) 23/126 (18%) 0.68(0.39e1.19)

0 — 5 32/250 (13%) 51/254 (20%) 0.64 (0.43e0.96) !0.10 14 (9e125)

Respiratorydistresssyndrome

4 3/75 (4%) 12/83 (14%) 0.38(0.13e1.07)

7 — 4 13/207 (6%) 21/196 (11%) 0.61 (0.32e1.19) !0.10 —

Grade III/IVintraventricularhemorrhage

5 1/137 (1%) 3/126 (2%) 0.50(0.08e2.96)

8 — 4 0/207 (0%) 4/196 (2%) 0.28 (0.05e1.64) !0.10 —

Necrotizingenterocolitis

5 3/137 (2%) 3/126 (2%) 0.84(0.19e3.77)

0 — 4 1/207 (1%) 2/196 (1%) 0.62 (0.08e4.67) !0.10 —

Neonatal sepsis 5 5/137 (4%) 13/126 (10%) 0.38(0.15e0.96)

0 16 (11e242) 4 8/207 (4%) 17/196 (9%) 0.47 (0.21e1.05) !0.10 —

Bronchopulmonarydysplasia

3 5/113 (4%) 6/90 (7%) 0.61(0.20e1.83)

0 — 1 7/135 (5%) 6/127 (5%) 1.10 (0.38e3.18) ! 0.10 —

Compositeneonatalmorbidityb

4 3/75 (4%) 16/83 (19%) 0.29(0.11e0.81)

0 7 (6e27) 4 17/207 (8%) 28/196 (14%) 0.60 (0.34e1.06) ! 0.10 —

Conde-Agudelo. Vaginal progesterone vs cerclage in women with a singleton gestation, previous spontaneous preterm birth, and a short cervix. Am J Obstet Gynecol 2018. (continued)

Report

ofMajo

rIm

pact

ajog.org

16American

JournalofObstetrics

&Gynecology

JULY2018

ALLGEMEINES

KRANKENHAUS




5 typische Konstellationen

ü Patientin mit 1 Frühgeburt in der AnamneseüProgesteron

ü Patientin mit 2 oder mehr Frühgeburten in der AnamneseüCerlage

ü Patientin ohne Frühgeburt und verkürzter CervixüProgesteron

ü Patientin mit einer Frühgeburt und verkürzter CervixüProgesteron oder Cerclage

ü Patientin mit aufgebrauchter Cervix und Fruchtblase sichtbar

üCerclage

Documents

Progesteron und Cerclage–fürwen? · ALLGEMEINES KRANKENHAUS Die menschliche Größe DER STADT WIEN undfeto-maternale Medizin ABTEILUNG FÜR Progesteron und Cerclage–fürwen?