A prior history of preterm birth is known to be a high-risk factor for spontaneous preterm birth [1, 2]. The mechanism for this recurrence is still not yet understood; however, cervical insufficiency is thought to be a part of the cause, especially if prior preterm delivery occurred during the second trimester. Transvaginal cervical cerclage has been indicated as a treatment option for cervical insufficiency. Currently, cerclage placement based on obstetric history at 13–14 gestational weeks is referred to as history-indicated cerclage (also known as prophylactic cerclage), as opposed to ultrasound-indicated cerclage or physical examination-indicated cerclage [3, 4, 5, 6, 7].

Although history-indicated cervical cerclage is considered for patients with cervical insufficiency, controversy remains regarding its use over serial ultrasound assessment. The widespread use of transvaginal ultrasound evaluation of cervical length has become more common to defer cerclage placement until cervical length is shortened. The question remains as to whether it is safe to undertake serial ultrasound assessment instead of history-indicated cerclage for women with a history of second trimester pregnancy loss related to painless cervical dilation. The risk of iatrogenic rupture of membranes may increase if cerclage is placed after cervical dilation [7].

In our institution, the decision of whether to perform history-indicated cerclage or serial ultrasound assessment (ultrasound-indicated cerclage if needed) is made based upon the request of the women, following Japanese guidelines [8]. The aim of this study was to compare the pregnancy outcomes between women who chose history-indicated cerclage and those who chose serial ultrasound assessment. We also assessed whether the previous mid-trimester pregnancy loss was due to a painless cervical dilation, i.e., whether or not it was truly due to cervical insufficiency.

In the study period, a total of 63 women met the study criteria. Of these 63 women, 28 received history-indicated cerclage and 35 received expectant management. Maternal characteristics are shown in Table 1. There were no significant differences between the two groups except for the number of suspicious cervical insufficiency cases.

The incidence of preterm birth was similar in both groups (Table 2). Of the women who had history-indicated cerclage, two (7.1%) delivered before 28 weeks, and one delivered at 29 weeks. Four (11.4%) of the women receiving expectant management delivered before 28 weeks. Among patients receiving expectant management, rescue cerclage was indicated due to cervical dilation at 17–23 weeks in 5/35 (14.3%) women.

Considering the aforementioned groups of those with cervical insufficiency or not, all preterm births occurred in Group A, i.e., women who had painless cervical dilation at prior pregnancy loss. No one needed rescue cerclage, and no preterm births before 34 weeks occurred in Group B. The incidence of preterm birth before 28 weeks in Group A was significantly lower in women who chose history-indicated cerclage than expectant management. Preterm birth rates before 34 weeks were also lower in the history-indicated cerclage group, although not significantly different (Table 3). Eight women who did not received history-indued cerclage in group A were monitored every 2 weeks for cervical length; all five patients who had a cervical length of less than 25 mm at 26 weeks underwent ultrasound-induced cerclage. Of the five patients who had ultrasound-induced cerclage, three delivered at full term, while two delivered less than 1 week after the cerclage. In the three cases where cervical length was maintained to 26 weeks, two women delivered at 26 and 27 weeks, respectively. Thus, all but one of the women who chose expectant management with a history of painless cervical dilation received intervention or underwent preterm delivery (Fig. 1).

Fig. 1.

Preterm birth rate in each management when divided into Group A (women who had painless cervical dilation at prior mid-trimester pregnancy loss) and Group B (women not in Group A; typically whose preceding mid-trimester pregnancy loss was caused by pPROM or sub-decidual hemorrhage). PTD, preterm delivery; pPROM, preterm premature rupture of the membranes.

We also collected information about obstetric complications that may be increased by cerclage. One cervical laceration and one emergent cesarean delivery due to dystocia occurred in the history-indicated cerclage group; however, there was no such case in the expectant management group. PROM occurred 6/28 (21.4%) in the history-indicated cerclage group and 5/35 (14.3%) in the expectant management group. There was no significant difference between the groups.

Our study revealed no significant differences in preterm birth rate between history-indicated cerclage and expectant management among women who had experienced mid-trimester pregnancy loss at 14–24 weeks, although history-indicated cerclage showed lower very preterm birth rates in the population who experienced mid-trimester delivery due to painless cervical dilation, typically, the membrane bridging.

A Cochrane review [10] analyzed 15 randomized controlled clinical trials. Women considered to be at high risk of preterm birth, justifying cerclage placement, were assigned to cerclage, alternative treatments, or no treatment. This analysis concluded that cervical cerclage reduced the risk of preterm births, but data were limited for all clinical groups. A recent population-based cohort study [11] showed that history-indicated cerclage placement in women with only single mid-trimester loss might relate to preterm delivery. In this study, 108 women received history-indicated cerclage among 2175 patients with a singleton pregnancy ending in birth or miscarriage from 14 to 28 weeks. Women with cerclage were significantly more likely to deliver at 37 weeks than those without (39.8% vs. 19.3%) it. This result is not compatible with our data. This divergence might arise from the difference in study participants, because the range of gestational age at the previous preterm birth of women was relatively broader, and there was no consideration given to the pattern of prior mid-trimester pregnancy loss in the cohort study.

Studies [12, 13, 14, 15] have been conducted to determine whether the placement of cervical cerclage according to the shortening of the cervical length (i.e., ultrasound-indicated cerclage) is an acceptable alternative to history-indicated cerclage. The incidence of preterm delivery has been reported to be similar in history-indicated cerclage groups and ultrasound-indicated cerclage groups. These findings can be interpreted to suggest that the ultrasound-indicated protocol reduces needless intervention, although this is still unclear. In the study by Simcox et al. [12], only 19% of women randomized to the elective cerclage group underwent intervention, while 32% of women randomized to the ultrasound-indicated group received cerclage.

According to the results of this study, history-indicated cerclage may be superior to expectant management for women with a history of cervical insufficiency if their history is precisely evaluated.

This study has some limitations. First, this study was a retrospective and small-size cohort study, and the number of women who chose expectant management in group A was small. Therefore, it may be difficult to generalize the results of this study. Next, there is a selection bias in whether or not to undergo history-indicated cerclage. A randomized control study for homogenous cervical insufficiency patients is warranted, but this may be difficult because women with cervical insufficiency have a wide variety of pathology and the prevalence of each is relatively low. Lastly, intrauterine infection is also an important cause of preterm birth. Although we had confirmed placental pathology in most previous mid-trimester pregnancy loss, we did not exclude histological infection cases in the cervical insufficiency group because we could not know whether the infection occurred after cervical enlargement, or the infection induced preterm labor. Microorganisms causing these types of infections probably migrate to the uterus before or early during pregnancy [1]. We did not take routine vaginal cultures over this study period; however, vaginal cultures for bacterial vaginosis during the first trimester and the administration of appropriate treatment might have improved obstetric outcomes.

Women with mid-trimester pregnancy loss experience anxiety regarding the possibility of another pregnancy loss. In Japan, similar to many other developed countries, the interval between pregnancies has increased due to advanced maternal age. Detailed evaluation of any preceding pregnancy losses and informed decisions regarding the indication of prophylactic cerclage should be undertaken.

In conclusion, our data suggests that the history-indicated cerclage can be indicated for women with a single history of mid-trimester pregnancy loss with painless cervical dilation. Serial transvaginal ultrasound surveillance may reduce unnecessary cerclage, there are cases in which delayed intervention leads to premature delivery. Further prospective studies are expected to determine which cases are appropriate for the history-induced cerclage.

The data that support the findings of this study are available from the corresponding author, AT, upon reasonable request.

AT, HT and TA designed the research study. AT and YI accumulated and analyzed the data. AT, HT, and TA wrote the manuscript. All authors contributed to editorial changes in the manuscript. All authors read and approved the final manuscript. All authors have participated sufficiently in the work and agreed to be accountable for all aspects of the work.

This study was approved by the Institutional Ethical Committee of the Japanese Red Cross Nagoya Daiichi Hospital (2019-038). Informed consent was waived due to the retrospective nature of this study, the Ethical Committee approved that. This study adheres to the Declaration of Helsinki.

This research was funded by Japanese Red Cross Aichi Medical Center Nagoya Daiichi Hospital Research Grant NFRCH 23-0024.

The authors declare no conflict of interest.