In monochorionic (MC) twin pregnancies, morbidity and mortality is 3–8 times higher than in singleton pregnancies (Litwinska et al., Reference Litwinska, Syngelaki, Cimpoca, Frei and Nicolaides2020). Excess morbidity and mortality are predominantly associated with unequally shared placental territories and the almost universal presence of vascular interfetal anastomoses between twins (Lewi et al., Reference Lewi, Cannie, Blickstein, Jani, Huber, Hecher, Dymarkowski, Gratacós, Lewi and Deprest2007).
Overall, one out of three MC twin pregnancies will develop specific complications due to twin-to-twin transfusion syndrome (TTTS), twin anemia polycythemia sequence (TAPS), selective fetal growth restriction (sFGR), and twin reversed arterial perfusion sequence (TRAP). Discordant anomalies (DA) are also more commonly seen in monochorionic pregnancies than in dichorionic pregnancies probably because of a teratogenic effect of the embryonic splitting or because of transfusion imbalances (Lewi, Reference Lewi2022).
These conditions may lead to fetal demise (FD) in one twin, causing death or brain injury in the surviving fetus due to severe exsanguination into the demised twin’s body and placental territory (Mackie et al., Reference Mackie, Rigby, Morris and Kilby2019). In some cases, selective fetal reduction may be a reasonable option to avoid loss or morbidity of the co-twin (Donepudi et al., Reference Donepudi, Hessami, Nassr, Espinoza, Sanz Cortes, Sun, Shirazi, Yinon, Belfort and Shamshirsaz2022).
In MC twins, selective reductions are performed using several techniques that aim to stop the intertwin exchange of blood immediately. Vaso-occlusive techniques such as bipolar cord coagulation (BCC), radiofrequency ablation (RFA), interstitial laser ablation (ILA) of the cord, and fetoscopy-guided cord coagulation with laser have been proposed for selective fetal reduction in complicated monochorionic twins (Gaerty et al., Reference Gaerty, Greer and Kumar2015). Associated complications of these techniques include premature rupture of membranes in up to 30% of the cases and preterm delivery. Perinatal survival rates vary between 65% and 92% depending on the technique and indication (van Hoek et al., Reference van Hoek, van Klink, Verweij, Middeldorp, Haak, Lopriore and Slaghekke2023).
The optimal therapeutic technique to achieve selective reduction in MC twins has not been clearly defined and is still conflicting (Yinon et al., Reference Yinon, Ashwal, Weisz, Chayen, Schiff and Lipitz2015). This may be due to the relatively small number of patients presented in each study, variation in procedure choice in different centers, variation in operator experience, and preference.
The aim of this study is to describe the perinatal outcomes of complicated monochorionic twin pregnancies after selective fetal reduction in a tertiary care center in southern Spain using BCC and ILA as vaso-occlusive techniques.
Material and Methods
Study Design and Population
This retrospective cohort study included all monochorionic twin pregnancies that required selective fetal termination in a referral center in fetal medicine in the South of Spain, the Maternal-Fetal Medicine Unit at San Cecilio University Hospital, between January 2008 and December 2023. This study was approved by the institutional ethics and research committee (reference number: 1416-N-20). The primary outcome was the survival of the co-twin, while secondary outcomes included gestational age at delivery, the incidence of PPROM, birth weight, and long-term neurodevelopmental outcomes.
Throughout the entire study period, all procedures were performed by a consistent team comprising one senior operator who was always present alongside two junior operators. This structured arrangement ensured continuous expert supervision and effective knowledge transfer, thereby minimizing any potential bias due to a learning curve and ensuring the uniform application of both techniques.
Data related to the procedure were collected from the patients and newborns’ medical records, including maternal age, parity, amnionicity, diagnosis, selective termination method, gestational age at the time of the procedure and perinatal outcomes of the surviving twins related to gestational age and birth weight at the delivery and PPROM. When postnatal outcome information could not be retrieved, follow-up phone calls to the patients or their referring physicians were made to obtain this information.
Ultrasound Assessment, Diagnosis and Procedure Indication
The indications for selective fetal reduction were severe cases of previable selective fetal growth restriction (sFGR) type II/III (discrepancy between the fetuses greater than 40%, persistent or intermittent absent/reversed end-diastolic flow in the umbilical artery or abnormal ductus venosus [DV] in the growth restricted fetus), twin reversed arterial perfusion (TRAP) sequence, discordant congenital anomaly, and twin-to-twin transfusion syndrome (TTTS) types III-IV where laser ablation of the placental anastomosis could not be technically performed.
Neuro-Developmental Outcomes
Long-term neurodevelopmental outcomes were analyzed by collecting data from medical records from the delivery date up to the present (mean follow-up period: 8.4 years), including information recorded by the healthcare staff and imaging studies (MRI, neuro-ultrasound). In our hospital, a specialized multidisciplinary pediatric unit oversees the management and care of preterm newborns delivered before 32 weeks until the age of 2 years. For neonates born after 32 weeks gestation, neonatal and childcare is provided by pediatricians who record routine follow-up; if any deviation from normal development is detected, the child is referred to a specialized multidisciplinary pediatric unit. In our unit, routine MRI is not performed after selective reduction unless abnormal neuro-sonographic findings are identified on ultrasound. No standard neurodevelopmental scoring system was used.
Surgical Procedures
Interstitial laser was performed following prophylactic antibiotic and tocolytics treatment using an 18G needle (Smiths Medical International, Ltd., UK) under local anesthesia and continuous ultrasound guidance. We used a 600 µm laser fiber inserted into the needle until it reached a few millimeters in front of the needle tip. Laser coagulation was performed using a diode laser at 20 W (Intermedic, Barcelona, Spain) as it was described before (Scheier & Molina, Reference Scheier and Molina2011).
In cases of bipolar cord coagulation, the procedure was performed under local anesthesia and ultrasound-guidance, introducing a 3 mm bipolar electrocautery forceps through a 10-F sheath (Performer, Cook Medical, Bloomington, Indiana) disposable trocar inserted into the amniotic sac of the targeted fetus. The umbilical cord was grasped, and coagulation was performed for 10 seconds at a power setting of 10 W less than gestational age, which was gradually increased by 10 W in each cycle until bubbles appear on ultrasound. The success criteria were the absence of flow in the umbilical cord and cessation of fetal heartbeat of the selected twin.
Statistical Analysis
For the results, categorical variables were described using absolute and relative frequencies, whereas continuous variables were described using means and standard deviations or medians and interquartile ranges, depending on the data distribution. The Shapiro-Wilk test was used to assess normality. For comparisons, the chi-square test or Fisher’s exact test was used for categorical variables, and the Student’s t-test or Wilcoxon rank-sum test was employed for continuous variables, as appropriate. All statistical analyses were performed using R software version 4.3 (R Development Core Team, 2010).
Results
A total of 84 selective terminations procedures in monochorionic twin pregnancies were performed in our unit from January 2008 to December 2023, 30 with ILA and 54 with BCC. No clear trend in the use of these techniques was identified over the period studied.
Maternal and pregnancy characteristics of the women undergoing selective fetal termination are detailed in Table 1, revealing no notable differences between the ILA and BCC groups overall. However, the mean gestational age at which the procedure was performed was significantly lower in the ILA group (mean difference: -5.28 weeks; 95% CI [-6.52, -4.04]; p < .001). Notably, within our cohort, there were nine cases of monochorionic and monoamniotic twins, with seven cases managed using BCC and cord transection, while only two TRAP cases were treated with ILA.
Table 1. Characteristics of pregnant women undergoing selective fetal termination
Note:
* Mean (SD); n (%).
TRAP sequence was the indication in all the patients treated with ILA (30/30). Among patients treated with BCC, the most common indication was sFGR, occurring in 31 of 54 cases (57%), followed by discordant anomalies with 15 cases (28%), TRAP sequence 6 cases (11%) and TTTS 2 cases (3.7%) (Figure 1).
Figure 1. Indications for selective fetal termination.
Perinatal outcomes after selective fetal reduction according to the type of technique used are shown in Table 2. The survival rate was higher among those treated with BCC compared to those treated with ILA (87% vs. 67%, p = .026). The proportion of fetal death occurring before 24 weeks was greater in the ILA group (30%) compared to the BCC group (7.4%; p = .010). Additionally, gestational age at delivery was significantly lower in those treated with BCC (p = .021), and preterm delivery was more common in BCC (50%) than in ILA (14%) (p = .005). Lastly, birth weight was lower in BCC (median 2595 g) compared to ILA (median 3030 g), with a p value of .050, indicating marginal significance.
Table 2. Perinatal outcomes after selective fetal reduction by technique
Note:
* n (%); median (IQR);
** Pearson’s chi-squared test, Fisher’s exact test, Wilcoxon rank sum test; PPROM, premature preterm rupture of membrane.
Regarding long-term neurodevelopmental outcomes, medical records based on pediatric evaluations and, in some cases, imaging studies, demonstrated normal developmental outcomes in 66 of the 67 co-twins followed. Only one case was described as having epileptic encephalopathy of unknown cause, accompanied by severe developmental delay, psychomotor retardation, and conduct disorder. This case was a monochorionic diamniotic twin pregnancy complicated with a discordant anomaly, a Dandy Walker malformation, with a normal genetic test result, treated with BCC at 18.1 weeks. At 34.5 weeks, the pregnancy ended by an emergency C-section due to severe maternal chronic hypertension. At the time of the study, the surviving co-twin was 15 years old. A perinatal ischemic was proposed as the primary cause of the diagnosis.
Discussion
Main Findings
In this cohort study of 84 cases, we evaluated and described the principal outcomes of selective fetal reduction in complicated monochorionic twins. Starting with the overall survival rate of the co-twin of 80%, the survival rate was higher after BCC (87%) than after ILA (67%). Our results have shown that ILA was performed earlier during pregnancy as this technique is minimally invasive compared with BCC. The results have also shown that after ILA the risk of preterm birth is lower than BCC, and this translates into improvements in gestational age at delivery as well as newborn birth weight.
Comparison with Previous Studies
Survival rate reported in the literature after selective termination in monochorionic twins varies from 60–85%. This was confirmed in a recent systematic review of selective fetal reduction with 97 complicated monochorionic twin pregnancies where the authors likewise compared BCC with ILA. They observed that the fetal survival rate was 82.3% (51/62) with BCC and 57.1% (20/35) after ILA, showing a lower survival rate after the application of the ILA technique (Donepudi et al., Reference Donepudi, Hessami, Nassr, Espinoza, Sanz Cortes, Sun, Shirazi, Yinon, Belfort and Shamshirsaz2022). Along the same line, van den Bos et al. (Reference van den Bos, van Klink, Middeldorp, Klumper, Oepkes and Lopriore2013) reported that in 131 cases where different fetal reduction procedures were applied and compared, they also observed lower survival rate in the ILA-treated group (46.7%) compared with the BCC group (77.8%).
In this study, fetal death before 24 weeks has been found higher in ILA (30%) than BCC (7.4%). The question to be resolved is whether the worst outcome regarding fetal death is related only to the gestational age at which the procedure was performed, or the specific technique applied as well as the indication for the procedure. It is important to mention that when the percutaneous technique is used (ILA), cord occlusion of the affected fetus is more uncontrolled, and it may result in the exsanguination of the healthy fetus on the affected one or on its placenta. In a systematic review by Rossi and D’Addario (Reference Rossi and D’Addario2009), the authors suggested that the association between gestational age at surgery and survival rate may depend on the technique performed to achieve selective feticide rather than the indication for surgery. When survival rate is stratified for indication to selective feticide, it is similar between pregnancies complicated with TTTS (82%), severe malformations (85%), and TRAP sequence (75%) performing the same procedure at the same gestational age (Rossi & D’Addario, Reference Rossi and D’Addario2009).
In our study, the mean gestational age at delivery was 36.6 weeks in BCC (33.1−38.4 weeks) and 38.6 weeks in ILA (37.1−39.8 weeks). Our results have also shown that BCC is associated with a higher percentage of preterm delivery (50%) when compared with ILA (14%). In addition, in our series, the median birth weight was lower in BCC 2.550 grams (1820–3000g) than in ILA 3.030 grams (2700–3220) with marginal significant differences, probably associated with a lower gestational age at birth. Contrary to what is described in a systematic review by Donepudi et al. (Reference Donepudi, Hessami, Nassr, Espinoza, Sanz Cortes, Sun, Shirazi, Yinon, Belfort and Shamshirsaz2022), the mean gestational age at delivery was not significantly different between BCC (33.61 GA) and ILA (33.06 GA). These distinctions with the literature may be related to the smaller sample size of the series as well as the different techniques used and a mean GA at delivery where liveborn and miscarriages were included. In another study published by van Hoek et al. (Reference van Hoek, van Klink, Verweij, Middeldorp, Haak, Lopriore and Slaghekke2023), a mean GA at delivery in liveborn, excluding miscarriages, was described. The authors describe, more akin to our series, that in BCC the mean GA at delivery was 34.4 (+4.5) weeks, slightly lower than in ILA 35.3 (+5.1) weeks. In addition, the birth weight in the van Hoek et al. study was also inferior in BCC (2331 grams) than in ILA, 2585 grams.
In our series, TRAP sequence was the only indication in which both ILA and BCC were achieved. Late referrals determined the decision of performing BCC when the gestational age was above 16–18 weeks and a proper umbilical cord was present in the acardiac fetus. On the other hand, when the diagnosis was done during the first trimester scan, ILA was selected. Although lower survival rates are described when ILA is operated compared with BCC, delays or no intervention between the diagnosis of TRAP sequence at 11–13 weeks’ gestation until 16–18 weeks is associated with spontaneous death of the pump twin in 33% of cases (Lewi et al., Reference Lewi, Valencia, Gonzalez, Deprest and Nicolaides2010). Further delaying interventions can determine spontaneous cessation of flow in the acardiac twin in 60% of the cases and the mortality rates for the pump twins have been noted to be as high as 50−70%, as well as the possibility to suffer brain damage (Chaveeva et al., Reference Chaveeva, Poon, Sotiriadis, Kosinski and Nicolaides2014).
In our study, PPROM was higher in BCC (24%) than ILA (10%), supported by different published meta-analyses. In a systematic review by Rossi and D’Addario (Reference Rossi and D’Addario2009), the author describes the incidence of PPROM in relation with surgical technique, describing 21% of PPROM after BCC compared with 17% using other techniques. Donepudi et al. (Reference Donepudi, Hessami, Nassr, Espinoza, Sanz Cortes, Sun, Shirazi, Yinon, Belfort and Shamshirsaz2022) has also upheld the finding of greater PPROM and also PTB rates with BCC.
There were no cases of neonatal deaths. However, we report an isolated case of severe morbidity due to sequel encephalopathy of unknown cause versus epileptic encephalopathy after a preterm delivery at 34.5 weeks due to severe maternal hypertension, where a perinatal ischemic was proposed as the primary cause of the diagnosis. Lanna et al. (Reference Lanna, Rustico, Dell’Avanzo, Schena, Faiola, Consonni, Righini, Scelsa and Ferrazzi2012) reported normal neurological development in 82 out of 84 infants alive at 1–9 years’ follow-up. However, larger studies and a greater randomized number of series with long-term follow-up are needed. It is important to highlight that our study could analyze the long-term neurodevelopmental outcomes of 67 cases with an age range period from 1 to 17 years.
Strengths and Limitations
The main limitation of our study is that ILA was exclusively achieved for TRAP sequence cases diagnosed at the first trimester, while BCC was performed in all other indications. This limitation made it impossible to clarify whether the lower survival rate in the ILA group is predominantly related to the technique applied, the indication, or the gestational age at the time of the intervention. However, it has been published that one out of three pump-twins complicated with TRAP sequence dies from delaying the intervention until 16–18 weeks where BCC is possible to achieve (Lewi et al., Reference Lewi, Valencia, Gonzalez, Deprest and Nicolaides2010), and this justifies our preferences for ILA technique at the early first trimester scan with a 67% survival rate of the pump-twin. The lack of an objective method of neurodevelopmental assessment could be another limitation when it comes to analyzing these outcomes.
On the other hand, our database belongs to a single center, which makes the information reliable in terms of homogeneity. Second, our study included a higher number of cases compared with other published series, which makes it suitable to be use as a part of future meta-analysis to elucidate the best technique and gestational age at the time of a fetal selective termination in complicated monochorionic twins depending on the specific indication. The long-term follow-up of these children, either in multidisciplinary pediatric units or by general pediatricians, is one of the strengths of this case series.
Based on our experience, BCC seems to have a better overall survival than ILA in complicated monochorionic twins selective terminations. However, we could not clarify whether this difference was due to the technique itself or the different gestational age at the time of the procedure as well as the specific indication.
Acknowledgment
The authors would like to acknowledge the valuable feedback provided by colleagues in the Department of Obstetrics and Gynecology during the preparation of this manuscript.
Financial support
This research received no specific grant from any funding agency, commercial or not-for-profit sectors.
Competing interests
The authors confirm that there are no conflicts of interest. This work has not been funded by any public or private entity.
Ethical standards
The authors assert that all procedures contributing to this work comply with the ethical standards of the relevant national and institutional committees on human experimentation and with the Helsinki Declaration of 1975, as revised in 2008.
