Physical examination-indicated cerclage in twin pregnancies compared with singleton pregnancies

Abstract Objective The aim of this study was to compare pregnancy outcomes of physical examination-indicated cerclage in twin and singleton pregnancies with bulging membranes. Methods All women with bulging membranes in the second trimester of pregnancy who were admitted to La Fe University and Polytechnic Hospital from January 2009 to January 2022 were included. A total of 128 participants were enrolled, 102 singleton pregnancies and 26 twin pregnancies. All patients underwent an amniocentesis to rule out intra-amniotic inflammation (IL-6 < 2.6 ng/mL). Cerclage was placed in the absence of intra-amniotic inflammation. Results Compared with singleton gestations, twin pregnancies displayed a significantly higher prevalence of nulliparity and assisted reproductive techniques. The incidence of intra-amniotic inflammation/infection was similar in both groups (68.62% in singleton vs. 65.38% in twin pregnancies). The average gestational age of delivery without cerclage in singleton gestations was 23.83 weeks (95% CI 22.82–24.84) and in twin pregnancies, it was 23.69 weeks (95% CI 21.8–25.57). The average gestational age at delivery among patients with cerclage was 37.27 weeks (95% CI 35.35–39.19) in singleton gestations and 36 weeks (95% CI 33.51–38.63) in twin pregnancies, with no significant differences. Time from diagnosis to delivery in patients with IL-6 < 2.6 ng/mL was 79.88 days, and in those with IL > 2.6 ng/mL was 10.87 days. Gestational age at delivery was significantly higher in both singleton and twin pregnancies with cerclage, compared with those without cerclage (log-rank p < .001). Conclusions Singleton and twin pregnancies with bulging membranes behave similarly when cerclage is placed in the absence of intraamniotic inflammation/infection.


Introduction
Preterm birth (PTB) affects 5-18% of pregnancies and is a leading cause of neonatal death [1].Thus, screening for the risk of PTB is essential in twin [2] and singleton pregnancies [3,4].A recent study has analyzed the intrauterine and extrauterine variables related to prematurity with statistically significant evidence.Placenta praevia, pregnancy-induced hypertension, antibiotics, cervical length, physical exercise, fetal growth, maternal anxiety, preeclampsia, and antihypertensives increase the risk of prematurity.Thus, an accurate estimation of the risk of PTB can be performed with these factors [5].
Cervical insufficiency is a well-known risk factor for PTB in both singleton and twin pregnancies [6].It is defined as the inability of the uterine cervix to retain a pregnancy during the second trimester [7].Twin pregnancies, which have increased dramatically in the latest decades [8] due mainly to assisted-reproductive technology and advanced maternal age at the time of conception, are associated with an increased risk of PTB [9].
Although the etiology of cervical insufficiency is unclear, data suggest an association with subclinical inflammation and infection.Actually, it has been reported that 52% of women admitted between 14 and 24 weeks of gestation with cervical insufficiency have a microbial invasion of the amniotic cavity [10].
When cervical dilatation and exposed membranes in the second trimester are demonstrated, treatment options include expectant management or placement of a physical examination-indicated cerclage.Physical examination-indicated cerclage is associated with a significant improvement in neonatal survival and prolongation of pregnancy in singleton pregnancies [11][12][13].Data are scarce regarding the usefulness of physical examination-indicated cerclage in twin pregnancies with bulging membranes.Thus, the performance of an emergent cerclage in twin pregnancies is not robustly indicated [14].One study recently published states that prematurity and perinatal mortality are reduced in patients with twin pregnancies and cervical incompetence who undergo cerclage [15].Moreover, another recent study elegantly shows that physical examination-indicated cerclage in twin pregnancies prolongs latency to delivery similarly to singleton gestations [16].Thus, the aim of this study was to compare pregnancy outcomes of physical examination-indicated cerclage between twin and singleton pregnancies with premature cervical dilatation and exposed fetal membranes in the vagina.

Study design and participants
All pregnant women with bulging membranes in the second trimester of pregnancy who were admitted to La Fe University and Polytechnic Hospital in Valencia, Spain, from January 2009 to January 2022 were included.A total of 128 participants were enrolled in the study, 102 singleton gestations and 26 twin pregnancies (Figure 1 [17]).The McDonald cerclage technique was performed in 16 patients (10 singleton and six twin pregnancies).The Shirodkar type was placed in 23 patients (20 singleton and three twin gestations).Exclusion criteria were uterine contractions, clinical chorioamnionitis [18], premature rupture of membranes, active bleeding, cervical dilatation >3 cm, congenital fetal anomaly, fetal death, and >26 weeks' gestation [19].Characteristics of singleton and twin pregnancies are shown in Table 1 and Figure 1.
The study was approved by La Fe University Hospital Ethics Committee.Informed consent was obtained from all participating women.
The presence of bulging membranes is defined as the presence of cervical changes of the internal os (e.g. 1 cm dilated or prolapsed membranes) detected on physical examination [20].
According to the hospital's protocol [21], patients with bulging membranes received physical examination as well as vaginal and abdominal ultrasonography at the time of their admission.Blood samples of pregnant patients were obtained, and electronic fetal monitoring was performed.All patients underwent an amniocentesis (in the case of twin pregnancies, only the first amniotic sac was punctured) with a 20-gauge 15-cm length needle and a 10-cm 3 aspiration syringe.Within the hospital's protocol, 70 cm 3 of amniotic fluid were extracted, 10 cm 3 for microbial analysis, 10 cm 3  for biochemical analysis (IL-6, leukocytes, glucose, leukocyte esterase, LDH, and procalcitonin), and additional 50 cm 3 for amnioreduction.Amnioreduction could benefit those patients who subsequently may take advantage of a cervical rescue cerclage [22][23][24].Intraamniotic inflammation was defined by amniotic IL-6 >2.6 ng/mL [25].Intraamniotic infection was determined by positive bacterial culture [19].Cerclage was placed in the absence of intraamniotic inflammation or infection.Moreover, cerclage was placed if subclinical chorioamnionitis was ruled out.The following criteria defined subclinical chorioamnionitis: amniotic glucose <15 mg/dL, amniotic leucocytes >50/mm 3 , positive amniotic leucocyte esterase, and amniotic LDH >410 U/mL [26].Finally, patients with clinical chorioamnionitis did not receive a cerclage.Clinical chorioamnionitis was defined by the following criteria: maternal temperature (>38 C) and one of these three criteria: fetal tachycardia (>160 bpm during >10 min), serum leukocytosis (>15,000/mm 3 ) or purulent cervical discharge [27].

Assays
Levels of IL-6 in the amniotic fluid were measured with commercially available lateral flow immunoassay designed for the quantitative measurement of human IL-6.Interpretation was carried out by Milenia POCScan Reader or PicoScan Systems (MileniaBiotec, Gieben, Germany).The laboratory results were shown within 90 minutes after the extraction [25].

Cervical cerclage placement
All patients with bulging membranes beyond the external cervical os without intraamniotic inflammation or infection, or subclinical or clinical chorioamnionitis, were proposed to receive a rescue cervical cerclage.Those patients who rejected the invasive procedure or had intraamniotic inflammation/infection or subclinical or clinical chorioamnionitis were treated with bed rest with low-molecular-weight heparin 40 mg subcutaneously (daily) and intravenous antibiotics (ampicillin 1 g every six hours combined with gentamycin 240 mg daily for five days).
McDonald or Shirodkar cerclage techniques were used according to the surgeon's decision at the moment of the procedure.Under spinal anesthesia, prolapsed fetal membranes were replaced with an inflated Foley catheter (14 F, 30-cm 3 balloon; Yushin Medical Co., Seoul, South Korea), and then cerclage was performed using 5-mm Mersilene tape (Ethicon Inc., Auneau, France).Ten trained practitioners carry the procedure out at the tertiary hospital.Before surgery, a 100 mg suppository of indomethacin is administered, and then treatment is maintained for 48 h with 25 mg indomethacin every 8 h.Regarding the type of suture, a study published in 2016 [28] states that braided suture induces a persistent shift toward dysbiosis of the vaginal microbiome characterized by reduction of Lactobacillus spp.and enrichment of pathogens causing premature cervical remodeling.We decided to continue using Mersilene suture as a subsequently published study [29] has shown no significant differences in maternal infection rates and adverse neonatal outcomes.There is also no difference in the prolongation of pregnancy when comparing the different suture materials used for the indicated locks.

Statistical analysis
The comparative study used non-parametric tests, the Mann-Whitney U-test for quantitative variables and the Kruskal-Wallis test for categorical variables.
The Kaplan-Meier curve was used for the survival analysis, and log-rank tests were applicated for the Kaplan-Meier curve comparisons.SPSS 20 (IBM, Chicago, IL) was used for statistical analysis.A p value <.05 was considered significant.

Results
Twin pregnancies displayed a significantly higher prevalence of nulliparity and assisted reproductive techniques than singleton gestations.No significant differences were found regarding the mean gestational age at diagnosis of bulging membranes and the interval from diagnosis to delivery.Cerclage was placed in approximately one-third of the patients with singleton and twin pregnancies.No complications during the cerclage or in the postoperative period were observed.
The mean gestation length from diagnosis to delivery in cerclage cases was 11 weeks in singleton pregnancies and 12 weeks in twin pregnancies, with no significant differences in any of the analyzed values.
Two patients with a singleton gestation and IL-6 <2.6 ng/mL did not consent to receive a cerclage.Thus, they received expectant management.One of the patients was diagnosed on day 178 and delivered on day 262.The interval of days from diagnosis to delivery was 84.She spent 69 days at the hospital and had a spontaneous term birth.The newborn's birth weight was 2840 g.The second patient was diagnosed on day 185 and delivered on day 259.The interval from diagnosis to delivery was 74 days.She was admitted to the hospital for 25 days and had a spontaneous term birth.The newborn's birth weight was 2400 g.
The mean birth weight in singleton pregnancies with cerclage was 2350 g, whereas in those without cerclage was 1012 g.On the other hand, the mean birth weight of the first and second twin in pregnancies with cerclage was 1934 g and 1922 g, respectively, while in pregnancies without cerclage was 675 g and 662 g, respectively, with significant differences (p < .01).
No significant differences in the analyzed amniotic fluid parameters were found in singleton and twin gestations (Table 2).The incidence of IL-6 > 2.6 ng/mL was similar in singleton and twin pregnancies, and positive culture was obtained in less than half of the patients.
A total of 28 positive cultures were shown in singleton pregnancies.The most frequent isolated bacteria were Gram þ (Streptococcus and Staphylococcus), which appeared in 14 patients (50%), followed by Gram -(E.coli, Klebsiella, Fusobacterium, etc.), which were detected in seven cases (25%), Ureaplasma was isolated in five cases (17.85%), and Candida albicans was present in two patients (7.14%).Concerning the eight twin pregnancies with a positive amniotic fluid culture, in four cases Gram þ bacteria were detected (50%), in two patients Grambacteria were isolated (25%), and in two women Ureaplasma was shown (25%).
The gestational age at delivery was significantly higher in both singleton and twin pregnancies with the cerclage, compared with those without the cerclage (log-rank p < .001)(Figure 2).Interestingly, the gestational age at delivery was not significantly different in singleton and twin pregnancies with cerclage, and singleton and twin gestations without cerclage (Figure 3 and Table 3 show the perinatal outcomes of patients with bulging membranes according to IL-6 in amniotic fluid).No significant differences were seen regarding the time from diagnosis to delivery in singleton and twin gestations with and without emergency cerclage (Figure 4).The average gestational age at delivery without emergency cerclage was not significantly different in singleton pregnancies (166.8 days [95% CI 159.75-173.88])compared with twin gestations (165.8 days [95% CI 152.59-179.05]).When emergency cerclage was placed, no differences were also found regarding the gestational age at delivery in singleton gestations (260.9 days [95% CI 247.47-274.37])compared with twin pregnancies (252.5 days [95% CI 234.62-270.42]).

Interpretation
Emergency cervical cerclage has been described as a treatment option to prolong pregnancy and reduce prematurity in singleton and twin pregnancies [11,20,30].Amniocentesis before the rescue cerclage is  still controversial.However, several authors suggest it because the diagnosis of subclinical chorioamnionitis significantly worsens the pregnancy outcome [21,31].
As this study has shown, ruling out intraamniotic inflammation and infection in patients with bulging membranes in the second trimester of pregnancy is essential to improve perinatal outcomes before placing an emergency cerclage.
Recently, two meta-analyses have demonstrated that emergency cerclage is associated with a significant improvement in perinatal outcomes in both twin and singleton pregnancies with a dilated cervix [12].In addition, for twin pregnancies with a cervical length <15 mm, cervical cerclage is associated with a significant reduction of PTB [32].
In twin pregnancies, the rate of cervical insufficiency is 5.0%, which is significantly greater than in singleton pregnancies (0.05-1.8%) [3].Accordingly, in this study, cervical insufficiency in twin gestations appears earlier than in singleton gestations, and it is associated with higher rates of nulliparity and assisted reproductive technologies.
Interleukin-6 is the major proinflammatory cytokine released in the amniotic fluid in response to bacterial invasion [33].The production of cytokines stimulates the synthesis of prostaglandins, which induce cervical ripening and its subsequent dilatation [5].

Main findings
In this study, amniocentesis of the presenting twin and in the singleton pregnancy to rule out intraamniotic inflammation or infection was performed.Cerclage was placed only if IL-6 was 2.6 ng/mL  (around 65% in both types of gestation).No differences were found in the amniotic fluid parameters between twin and singleton gestations.A total of 27.4% of cultures were positive in singleton gestations with bulging membranes, compared to 30.76% in twin gestations.The incidence of positive culture among singleton and twin pregnancies with bulging membranes is lower than that reported by Romero et al. [10].These authors described a prevalence of microbial invasion of the amniotic cavity of 51.5% in singleton pregnancies.The differences may be due to socioeconomic and ethnic dissimilarities between study populations.Whereas Romero et al. [10] described that the most common microorganisms isolated in patients with cervical insufficiency were Ureaplasma urealyticum, Gardnerella vaginalis, Candida albicans, and Fusobacterium sp.[10], in this work, the most prevalent bacteria have been Gram-positive (Streptococcus and Staphylococcus), followed by Gramnegative (Fusobacterium sp., E. coli, etc.), Ureaplasma urealyticum, and Candida albicans.
The determination of IL-6 in the amniotic fluid to rule out intraamniotic inflammation improves the perinatal outcomes of patients with twin and singleton pregnancies who require a cerclage indicated by physical examination.It would be very interesting to avoid amniocentesis before the cerclage to reduce the possible complications of the invasive technique.A prior study of our group demonstrated that vaginal IL-6 correlates with intraamniotic IL-6 [34].Further studies could determine whether vaginal IL-6 could be useful in ruling out intraamniotic inflammation.

Strengths and limitations
The study's main limitation is the limited sample size, particularly for twin pregnancies.Nonetheless, the consistency with other studies suggests that the results are reliable.The addition of data regarding emergency cerclage in twin pregnancies strengthens this work.

Conclusions
In the present study, emergency cerclage is associated with advanced gestational age at delivery, higher birth weight, and a prolonged period from diagnosis to delivery in singleton and twin gestations, compared with those without the cerclage.Given that there is a lack of data regarding pregnancy outcomes of emergency cerclage in twin pregnancies, the present findings could be helpful to counsel patients with a twin gestation and bulging membranes in the second trimester of pregnancy, without intra-amniotic inflammation and infection, who could benefit from a physical examination-indicated cerclage.

Figure 1 .
Figure 1.Flowchart of the study design.

Figure 2 .
Figure 2. Gestational age at delivery in singleton and twin gestations with (continuous line) and without emergency cerclage (dotted line).

Figure 4 .
Figure 4. Time from diagnosis to delivery in singleton and twin pregnancies according to the placement of an emergency cerclage.Dotted line without cerclage and continuous line with cerclage.

Table 1 .
Basal characteristics and pregnancy outcomes in twin and singleton gestations.

Table 2 .
Amniotic fluid studied parameters in singleton and twin gestations with bulging membranes in the second trimester of pregnancy.

Table 3 .
Pregnancy outcomes in patients without intra-amniotic inflammation or infection and patients with it and patients with intra-amniotic infection.