The number of twin births has increased in recent years [1] . About 50% of these are born before 37 weeks, which implies important fetal morbidity and mortality. As prematurity can be considered a multifactorial syndrome, preventive measures and management of cases at risk usually include actions related to the main contributing factors for preterm labor. One of the several factors which increase preterm birth is the presence of short cervix. Shortening of the cervical length (CL) is a predictor of preterm birth in singleton and twin pregnancies. The risk of preterm delivery is inversely related to cervical length (CL) assessed at 22 - 26 weeks of gestation, and a CL ≤ 25 mm is associated with 28% risk for birth before 28 weeks [2] [3] . The cut-off measure to classify a cervix as short in the second trimester is controversial for twin pregnancies. Different studies describe cut-off measures of 38 mm, 30 mm or 25 mm [2] [4] [5] [6] . In our clinical practice, a CL below 30 mm in twin pregnancies is considered as short. The cervical length is not routinely measured in Brazil’s public health care system, and there is no national recommendation on this topic. Nevertheless, the medical staff responsible for the prenatal care of the patients described on this paper followed the International Society of Ultrasound in Obstetrics and Gynecology guidelines [7] . Currently, there are different strategies to manage twin pregnancies with short cervix (TPSC), most of them with no proven effectiveness. Cervical cerclage has been associated with increase in the frequency of adverse outcomes in twin pregnancies [8] . A systematic review and meta-analysis showed that vaginal progesterone reduced the rate of preterm delivery in TPSC [9] [10] . Likewise, a randomized study associated the use of cervical pessary in twin pregnancies of mothers with a short cervix with significant reduction of prematurity [9] . In other recent meta-analysis, intravaginal progesterone 400 mg/day was more effective than in the doses of 100 or 200 mg in cases of TPSC [11] .

In this case series we describe six twin pregnancies that were considered as high-risk for preterm labor due to short cervix (CL < 30 mm) at second trimester. The patients were selected from one of the authors’ private clinic (J.V.), after screening for short cervix on second trimester.

Cervical length was obtained between 20 and 24 weeks of pregnancy by transvaginal ultrasound, using a General Electric (GE) VOLUSON E8, with an endocavity transducer at a frequency of 7.5 MHz. Ultrasound was performed with the patient in a dorsal lithotomy position and with empty bladder. The cervix was measured along its longitudinal axis. The cervix occupied approximately 50% - 75% of the image and pressure from the probe on the cervix was as little as possible. Cervical measurement was obtained by placing calipers at the external and internal OS, using a straight line between both of them. If the cervix was curved and the straight-line cervical length measurement was short, measurement obtained in two or more segments was performed to provide a more accurate estimation. The examination lasted from 3 to 5 minutes to detect possible changes in cervical length, and at least 3 measurements were obtained along that period.

The patients were managed on an individual basis. Some of them had multiple risk factors for prematurity, mostly previous infections. All patients had a cervical pessary (Ingamed^Ò Brazil) inserted, received vaginal micronized progesterone at a dose of 400 mg/day and were removed from their work activities. The main baseline characteristics, CL, gestational age at pessary insertion and individual risk factors are described on Table 1.

The main maternal and fetal outcomes are described on Table 1. The mean CL and gestational age at pessary insertion were respectively 21 mm (range 10 - 28 mm) and 22 weeks (range 16 - 24 weeks). Mean age of delivery was 31 weeks

NICU: neonatal intensive care unit. GBS: Group B Streptococcal colonisation. *Term delivery. **Fetal growth restriction. ***Abnormal fetal Doppler. ****Delivey indicated by medical reasons.

(range 26 - 34), half of them due to spontaneous onset of labor. In three cases the decision to deliver was due to fetal conditions and not due to preterm labor: one because of non-reassuring well-being test and two because of intrauterine growth retardation associated to abnormal Doppler ultrasound. The mean length of pregnancy since pessary insertion to birth was 9 weeks (range 2 to 17 weeks).

All pregnancies were delivered by caesarean, and all patients received antenatal corticosteroids for fetal pulmonary maturation before delivery. Patients with a positive group B streptococcal colonisation test at the onset of preterm labor received antimicrobial prophylaxis. The neonatal mortality rate was 8%, and the newborns who deceased had important anatomical problems (gastroschisis).

The risk of spontaneous preterm birth increases significantly with the number of risk factors for prematurity [12] . In our group of patients, we observed that in the two cases where birth occurred before 28 weeks there were multiple factors related to prematurity such as polyhydramnios, hypothyroidism, first-trimester bleeding, bacterial vaginosis, urinary tract infection, amniotic fluid sludge, cervical funneling and/or fetal malformations [13] [14] [15] .

The prevention of premature birth should be carried out throughout the prenatal visits by screening all the possible associated conditions in order to provide proper and immediate treatment. The use of tocolytics, corticosteroids and antibiotic therapy may have contributed significantly to the success of the presented cases since, with the exception of fetuses with major malformations, all babies were discharged without serious sequelae.

The prevention of prematurity in twins is challenging, and the association of different strategies is frequently used. Nevertheless, it still needs to be studied more deeply. In order to significantly reduce prematurity incidence, we consider the CL measurement of fundamental importance in all twin pregnancies [7] [16] , as well as the use of progesterone at a dose of 400 mg/day in cases of short cervix [11] .

Different authors assert that cervical pessary may be useful to reduce neonatal morbidity and mortality in the subgroup of women with TPSC without any serious adverse effects [5] [6] . Nevertheless, the pessaries used on this case series were manufactured by Ingamed^Ò (“Ingamed-Materiais Médico Hospitalares” n.d.), which may differ from the Arabin^Ò pessaries (“ARABIN^Ò Cerclage Pessar Perforiert” n.d.) described on most studies [17] [18] . The Arabin^Ò pessary is not easily available in Brazil, though. The lack of a control group without pessary insertion does not allow drawing any conclusion about its efficiency from this study, but it is important to highlight that this is the first publication utilizing the Ingamed^Ò Brazilian pessary for preterm delivery prevention.

Considering the high morbidity and mortality associated to prematurity, especially when extreme, the reduction in the number of days of neonatal intensive care unit stay is important to reduce economic, family and social costs. Despite no official Brazilian recommendation for measurement of cervix in twin pregnancies, screening for asymptomatic cervical shortening in those patients seems to be necessary, once there probably are effective interventions that could be established. In this context, the pessary seems to be at least a safe measure to extend pregnancy length on selected TPSC.

The authors declare no conflicts of interest regarding the publication of this paper.

Vettorazzi, J., Salazar, C.C., Valério, E.G., Isotton, A.L., Petermann, A., Rosa, M.W. and Rostirolla, G.F. (2018) First Experience Using Brazilian Ingamed^® Cervical Pessary in Twin Pregnancy: A Case Series. Open Journal of Obstetrics and Gynecology, 8, 1282-1288. https://doi.org/10.4236/ojog.2018.813130