Key words
cerclage - preterm birth - progesterone - short cervix
Schlüsselwörter
Cerclage - Frühgeburt - Progesteron - Gebärmutterhalsverkürzung
Introduction
Preterm birth (PTB) remains a major cause of neonatal morbidity and mortality. Cervical
length, as measured by transvaginal ultrasound (TVUS), has been shown to be an accurate,
reproducible measure and an important predictor of PTB [1], [2], [3].
In the general obstetric population, cervical lengths are normally distributed and
when shortened, the risk of PTB significantly increases [1], [2], [3], [4]. An inverse relationship between cervical length and risk of PTB has been demonstrated,
where the likelihood ratio of PTB decreases linearly with increasing cervical length
above 20 mm [5].
Interventions to prevent PTB in the setting of risk factors or asymptomatic TVUS-detected
cervical shortening include cervical cerclage progesterone and cervical pessaries.
Cervical cerclage has been described and used from the 1950s [6], [7], [8]. Early results were conflicting due to difficulties in patient selection and in
predicting women who were at high risk of preterm labour [9], [10].
Progesterone, in different formulations, has also been used to prevent PTB and miscarriage
since the 1960s [11]. Vaginal progesterone was initially prophylactically administered to women with
historical risk factors for PTB. When compared to placebo, those treated with vaginal
progesterone had a significantly lower frequency of delivery prior to 34 weeks [12]. Vaginal progesterone has also been shown to improve outcomes in the setting of
TVUS-detected short cervix. In a large multicentre randomised control trial, the use
of vaginal progesterone in women with a cervical length < 15 mm significantly reduced
the incidence of PTB compared to placebo [13]. Cervical pessaries have also been shown to be an effective treatment for the prevention
of PTB in women with cervical shortening [14].
Interest has increased recently in the combined use of these interventions to prevent
PTB. A retrospective cohort study by Stricker et al. [15] did not show any benefit from the addition of vaginal progesterone to cervical pessary
for prevention of PTB. A similar finding was also seen in a randomised controlled
trial comparing vaginal progesterone and cervical pessary to vaginal progesterone
alone [16].
Our study is a retrospective examination of the potential benefit of adding vaginal
progesterone to cervical cerclage to prevent PTB in women with ongoing TVUS-confirmed
cervical shortening despite the presence of a cervical suture.
Materials and Methods
Study design
A retrospective case control study was performed of women with singleton pregnancies
who underwent cervical surveillance by the Preterm Labour Clinic at the Royal Womenʼs
Hospital (RWH), Melbourne, in the years 2004 – 2014. Ethics approval was obtained
from the RWH ethics committee.
Clinical records were used to identify women who underwent cervical cerclage alone
versus those who were treated with both cervical cerclage and vaginal progesterone.
Retrospective data including demographic data, cervical length measurements, interventions
and birth outcomes were collected from patient files and the clinical pathology viewer.
Patient selection and methods
Women referred to the Preterm Labour Clinic are identified as being at increased risk
of preterm labour based on a history of PTB, cervical surgery, uterine malformation
or TVUS findings of a shortened cervix (< 25 mm) on routine morphology scan.
The clinic protocol involves fortnightly visits from the 14th to 26th week of gestation
for TVUS cervical length measurements and cervical swabs for microscopy and culture.
During the research period, patients with a past history of documented cervical shortening
and treatment with rescue cerclage were offered elective placement of cervical cerclage.
Additionally, rescue cerclage or vaginal progesterone was offered to patients with
a TVUS finding of a shortened cervix (< 25 mm) according to individualised indications
for treatment and patient preference. Compliance with vaginal progesterone treatment
advice was not formally monitored as part of the pre-term labour clinic protocol.
Cervical cerclage was performed by a core small number of senior medical staff of
the PTL clinic who were accredited for the procedure, with the decision for McDonald
or Shirodkar cerclage being made according to clinician preference.
Following cerclage placement, patients continued to attend the pre-term labour clinic
for serial TVUS cervical length measurements. Adjuvant vaginal progesterone therapy,
at a dose of 200 mg daily, was initiated if there was evidence of progressive cervical
shortening of < 20 mm on subsequent TVUS cervical length measurements at the treating
clinicianʼs discretion. This became standard practice midway through the study period.
Analysis and results
The shortest TVUS measured cervical length of women in the two intervention groups
was recorded. Cases in each intervention group were matched according to exact shortest
cervical length measurements during the cervical surveillance period.
Each intervention group was compared with regard to birth outcomes ‘gestational age
at delivery’ and ‘birth weight’. Neonatal Apgar scores at 1 and 5 minutes along with
mortality and intensive care admissions were also assessed. The differences in outcomes
between the intervention groups were compared using paired t-tests. The data were
transferred to an Excel spread sheet (Microsoft Corp., Redmond, WA, USA) and analysed
using GraphPad Prism (GraphPad Software Inc., San Diego, CA, USA).
Results
A total of 66 women were identified; 33 women who underwent cervical cerclage alone
were matched to 33 women who received both cervical cerclage and vaginal progesterone.
Matching of cases was based on exact matching of the shortest recorded TVUS cervical
length measurement during the period of cervical surveillance (14 – 26 weeks of gestation)
with the average shortest cervical length measurements of both groups being 12.09 mm.
There was no difference in average maternal age, parity or smoking status between
the intervention groups, with effective pairing seen on paired t-testing.
The majority of women across both intervention groups underwent cervical surveillance
in the setting of previous PTB ([Table 1]).
Table 1 Baseline characteristics and results.
|
Group
|
p-value
|
|
Cerclage + progesterone (n = 33) mean
|
Cerclage alone (n = 33) mean
|
|
PTB: preterm birth
*– Elective cerclage: history-indicated, placed prior to ultrasound measured cervical
shortening.
**– Rescue cerclage: Ultrasound-indicated, placed in asymptomatic women diagnosed
with short cervix on ultrasound.
|
|
Age (years)
|
32 (23 – 41)
|
31 (22 – 42)
|
0.27
|
|
Current smoker
|
8
|
5
|
|
|
Indication for cervical surveillance
|
|
|
|
|
|
14
|
15
|
|
|
|
6
|
6
|
|
|
|
2
|
0
|
|
|
|
1
|
4
|
|
|
|
5
|
4
|
|
|
|
2
|
0
|
|
|
|
2
|
3
|
|
|
|
1
|
1
|
|
|
Parity
|
1.0
|
1.5
|
|
|
Indication for suture placement
|
|
|
|
|
|
9
|
10
|
|
|
|
24
|
23
|
|
|
Gestational age at suture placement (weeks)
|
17.5
|
17
|
|
|
Type of cervical cerclage
|
|
|
|
|
|
9
|
15
|
|
|
|
11
|
13
|
|
|
|
13
|
5
|
|
|
Gestational age at progesterone commencement (weeks)
|
19.5
|
–
|
|
|
Shortest cervical length (mm)
|
12.09
|
12.09
|
|
|
Gestational age at delivery
|
36.36
|
32.64
|
0.0036
|
|
Birth weight (g)
|
2830
|
2199
|
0.0065
|
|
Apgar score
|
|
|
|
|
|
7.6
|
7.0
|
0.25
|
|
|
8.4
|
7.6
|
0.18
|
|
Composite morbidity/mortality
|
14
|
17
|
|
|
|
0
|
3
|
|
|
|
1
|
3
|
|
|
|
7
|
7
|
|
|
|
6
|
4
|
|
The average age of gestation for cervical cerclage placement was also similar in both
intervention groups, at around 17 weeksʼ gestation. Vaginal progesterone administration
commenced on average at 19 weeks, reflecting the nature of the use of this intervention
as a further treatment for ongoing cervical shortening despite the presence of cervical
cerclage.
Delivery outcomes
The progesterone and cerclage group had an average gestational age at delivery of
36.36 (36 + 2) weeks with an average birth weight of 2830 g. The cerclage alone group
had an average gestational age at delivery of 32.64 (32 + 4) weeks and an average
birthweight of 2199 g. This difference was statistically significant (paired t-test,
p = 0.0036) This result was also reflected in the Kaplan-Meier curves, which show
the statistical difference in the percentage of ongoing pregnancies between the intervention
groups ([Fig. 1]).
Fig. 1 Kaplan-Meier pregnancy survival: statistical difference in ongoing pregnancies between
intervention groups.
Neonatal outcomes
No significant differences were found between the intervention groups when comparing
Apgar scores, perinatal mortality and intensive care admissions.
Discussion
The results of our study illustrate the potentially beneficial effects of adjuvant
vaginal progesterone in a cohort of women at high risk of PTB. The high risk for PTB
in this cohort is reflected by numerous risk factors. Firstly, the majority of patients
included in the study had a history of PTB. Moreover, the average shortest cervical
length of 12 mm and early cervical shortening, reflected by the average gestational
age at suture placement of 17 weeks, further contributed to the risk for PTB. However,
promisingly, the addition of vaginal progesterone in this high-risk cohort was seen
to prolong pregnancy and reduce early pre-term delivery.
The use of progesterone concurrently with cervical cerclage for the prevention of
pre-term birth has also been highlighted in a small number of studies.
The PREGNANT trial, which looked at the use of progesterone pessaries for the prevention
of PTB in women with TVUS cervical lengths of 10 – 20 mm, included 16 women who underwent
emergency cervical cerclage post randomisation, with 10 women in the vaginal progesterone
group and 6 in the placebo group. However, the small numbers of women who had adjuvant
cervical cerclage in each intervention group were not statistically significant; patients
were analysed in their randomised treatment groups on an intention-to-treat basis
with no sub-analysis performed [17]. When Owen et al. [18] looking at cervical cerclage for the prevention of PTB, they included a sub-analysis
of 39% of the study population that indicated an intention to use progesterone in
conjunction with the randomised intervention. The progesterone referred to in the
study was intra-muscular 17-alpha-hydroxy progesterone caproate, not vaginal progesterone.
A sub-analysis was done of the women who intended to use progesterone; however compliance
was not addressed. The analysis based on intention-to-treat found that there was no
interaction between cerclage and the intra-muscular 17-alpha-hydroxy progesterone
caproate.
As regards the concurrent use of vaginal progesterone and cervical cerclage, a study
by Jung et al. showed promising results for the prevention of PTB. This retrospective
cohort study looked at 53 women who underwent cervical cerclage after physical exam
indicated the need for cerclage. Eighteen of these women also received concurrent
vaginal progesterone from the time of cerclage placement. This was shown to be of
benefit for prolonging pregnancy [19]. However, without objective cervical length measurements in this study, it is difficult
to appreciate the risk for PTB in each intervention group and thus the true impact
of the use of concurrent vaginal progesterone and cervical cerclage in this cohort.
In addition, the rationale for the addition of vaginal progesterone in this study
is unclear. Without a clear indication for the concurrent use of vaginal progesterone
with cervical cerclage, it becomes difficult to appreciate any true benefit from its
addition and thus to justify any potential risks and the cost of its use. It may be
argued that vaginal progesterone is relatively safe and, given its established efficacy
in prevention of PTB, perhaps combining it with another proven intervention such as
cervical cerclage would result in greater efficacy of PTB prevention. Overall there
is limited long-term data on the use of vaginal progesterone and any adverse maternal
or foetal effects. There is some population data on the use of progestogens in infertility
linked to increased rates of certain cancers in the offspring [20]. Although this has not been demonstrated with the use of vaginal progesterone, maternal
side effects of discomfort and discharge and, importantly, the cost of administration
must also be taken into account. Thus, a clear indication for the addition of vaginal
progesterone needs to be established to justify its use as an additional intervention.
To the best of our knowledge this is the first study which has directly investigated
the role of vaginal progesterone as a rescue adjunct in women with cervical cerclage
and ongoing TVUS diagnosed cervical shortening.
In our study, vaginal progesterone was administered to women with TVUS evidence of
progressive cervical shortening despite cervical cerclage, in particular women with
cervical lengths < 20 mm, a length at which the risk for PTB is thought to increase
exponentially [5]. The increased risk of PTB with shortening cervical lengths has been shown in numerous
studies. In the absence of cervical cerclage, Iams et al. [4], using logistic regression analysis, showed that for each 1 mm increase in cervical
length the odds ratio of PTB was 0.91. Berghella et al. [21] similarly concluded that the risk of PTB decreased by 6% with every additional mm
in cervical length.
This increased risk of PTB has also been seen in women with progressive cervical shortening
despite cerclage. Drassinower et al. [22] showed that among women with cerclage, an accelerated rate of cervical shortening
is associated with spontaneous PTB whereas patients with cerclage who delivered at
term had a slower rate of cervical shortening. Sim et al. [23] demonstrated that there was a strong correlation between serial TVUS cervical length
measurements post cerclage and gestational age at delivery. Thus, ongoing cervical
surveillance of women with cervical cerclage has the potential to identify a cohort
of women who have progressive cervical shortening and remain at increased risk of
PTB.
International cervical screening guidelines suggest that there is insufficient data
to encourage continued routine cervical length surveillance post cervical cerclage
placement. This is thought to be primarily due to the lack of additional treatment
options for a short cervix post cerclage [24]. Indeed, when looking at the placement of a repeat reinforcing cerclage, results
have been mixed, even detrimental [25], [26]. Our study provides some preliminary evidence that the use of adjuvant vaginal progesterone,
used as a rescue intervention in women with cervical cerclage failure, may be of benefit
in the prevention of early pre-term birth.
It is important to note that the generalisability of our results is limited by the
small sample size and the retrospective case control study design. Interventions,
although largely dictated by the pre-term labour clinic protocol, were at times left
to clinician discretion and patient preference, and despite our positive findings,
a larger prospective investigation is required to confirm this potential benefit.
The intention of this study was a preliminary exploration of any potential benefits
of adjuvant vaginal progesterone for failing cervical cerclage; thus, comprehensive
data on neonatal morbidity or mortality was not addressed. Although the addition of
vaginal progesterone to cervical cerclage did prolong pregnancy and improve birth
weights, there was no difference in Apgar scores at the time of delivery. Given the
small sample size, significant conclusions cannot be drawn from the neonatal morbidity
and mortality data we presented, nor were neonatal outcomes followed up in the longer
term. As a result, it is important that future research into this topic includes further
evaluation of neonatal outcomes both in the long and short term, particularly as available
2-year follow-up data on the use of prenatal vaginal progesterone for the prevention
of PTB, albeit as a single intervention, suggests no long-term difference in neurological
outcomes when compared to placebo [27].
Despite the limitations, our study provides supportive evidence for the beneficial
effect of rescue adjuvant vaginal progesterone, in women with further cervical shortening
despite cervical cerclage, for prolonging pregnancy. This finding is in keeping with
the beneficial effects of vaginal progesterone seen when used concurrently with physical
exam-indicated cervical cerclage. Our preliminary findings provide clinicians with
some evidence for the therapeutic benefit of adding vaginal progesterone to failing
cerclage, a clinical situation that currently presents a challenge for management.
A randomised controlled trial would be useful to further investigate the role of vaginal
progesterone in the prevention of PTB in women with very short cervical lengths despite
cervical cerclage.
Conclusion
In summary, vaginal progesterone may improve outcomes in pregnancies with a short
cervix despite cervical cerclage, including prolonged pregnancy and higher birthweight,
compared to treatment with cerclage alone. This is the first study to specifically
examine the use of vaginal progesterone as a rescue adjunct to cervical cerclage in
women with TVUS evidence of ongoing cervical shortening despite cervical cerclage.
Furthermore, it supports recent published data on the potential benefit of progesterone
when used vaginally in combination with cervical cerclage. Further research in this
area is required to characterise and confirm the potential benefits of the concurrent
use of progesterone pessaries and cervical cerclage for the prevention of PTB and
to address the effects on neonatal outcomes.
