The aim of this study was to analyze perinatal outcomes after selective reduction in monochorionic pregnancies with the use of either radiofrequency ablation (RFA) or bipolar cord occlusion (BCO). This was a systematic review and metaanalysis that included all studies with ≥5 cases that described perinatal outcomes after BCO or RFA that were identified in PubMed, Embase, Web of Science, COCHRANE, CINAHL, Academic Search Premier, Science Direct, and MEDLINE that were published between 1965 and July 2014. For count data, incidence risk ratios (IRR; 95% confidence interval [CI]) were calculated with BCO as the reference standard. The analysis included 481 cases of BCO and 320 cases of RFA from 17 studies. The mean median gestations at procedure were 21.1 ± 1.2 weeks (BCO) and 18.8 ± 2.5 weeks (RFA; P = .03). The rate of cotwin death was higher in the RFA group (14.7%) vs the BCO group (10.6%; IRR, 1.38; 95% CI, 0.93–2.05; P = .11). The live birth rate was 81.3% for the RFA group and 86.7% in the BCO group (IRR, 0.93; 95% CI, 0.80–1.09; P = .41). BCO had higher neonatal death rates (8.1%) vs RFA (4.5%; IRR, 0.56; 95% CI, 0.30–1.04; P = .07). Overall survival was 76.8% for RFA and 79.1% for BCO (IRR, 0.97; 95% CI, 0.82–1.14; P = .72); however, none of these differences were statistically significant. Preterm premature rupture of membranes occurred in 17.7% of RFA cases and 28.2% of the BCO cases (IRR, 0.63; 95% CI, 0.43–0.91; P = .01). The mean median gestational age at delivery was 34.7 ± 1.7 weeks in the RFA group and 35.1 ± 1.6 weeks in the BCO group. Our data do not demonstrate clearly the superiority of 1 procedure over the other. The clinical situation and preference of the operator are important considerations. Rates of preterm delivery and preterm premature rupture of membranes remain substantial for both procedures.
Selective fetal reduction is a management option that sometimes is required in complicated monochorionic pregnancies. Prevention of an acute interfetal transfusion during selective fetal reduction requires ablation of blood flow in the umbilical cord. If complete interruption of blood flow is not achieved during the procedure, the presence of placental vascular anastomoses could result in an acute shift of blood away from the healthy fetus into its dying cotwin. The resulting severe hypotension during this agonal event is believed to be one of the main causes of cerebral injury (multicystic encephalomalacia, germinal matrix or parenchymal hemorrhage, or grey matter lesions) to the surviving twin and, if severe enough, can cause death. Currently, the 2 most commonly used techniques for selective reduction are bipolar cord occlusion (BCO) and, more recently, radiofrequency ablation (RFA). RFA involves generating changes in alternating current at very high frequencies (200-1200 KHz) between the tines of the RFA needle. Because the electrical current alternates in various directions between the tines, tissue ions become agitated as they attempt to align with the electrical field. Frictional heat is then produced that results in very high tissue temperature, which causes tissue coagulation and necrosis.
Both techniques are minimally invasive, can be performed under ultrasound guidance, and in experienced hands are effective and can achieve complete ablation of blood flow in the targeted fetus. However, the initial enthusiasm for RFA has been tempered by more recent reports that have suggested that its use was associated with a decrease in the overall survival rate, despite a lower risk of preterm premature rupture of membranes (PPROM) and preterm labor when compared with BCO.
To date, there has not been a randomized controlled trial that has compared these 2 techniques. We performed a systematic review and metaanalysis to focus on perinatal outcomes in monochorionic pregnancies after selective fetal reduction with the use of either RFA or BCO.
Material and Methods
Articles that were published between 1965 and July 2014 for the analysis were obtained after searches of online electronic databases for all relevant publications that described perinatal outcomes after selective reduction by either BCO or RFA in monochorionic twin pregnancies. Independent searches were carried out by the author and the librarian at the Mater Research Institute–University of Queensland. Databases that were searched included PubMed, Embase, Web of Science, COCHRANE, CINAHL, Academic Search Premier, Science Direct, and MEDLINE. The following MESH terms or keywords were used: bipolar cord coagulation ( BCO ) or occlusion or radiofrequency ablation ( RFA ); monochorionic pregnancy , monochorionic , or twin . A manual search of the reference lists of the primary articles was carried out to identify relevant articles that were not captured by the electronic searches. The search was limited to English-language articles.
Individual case reports or case series with ≤5 cases or articles that described techniques other than RFA or BCO, personal communications and reviews, conference abstracts, book chapters, and guidelines were excluded from the analysis. Articles were also excluded if either technique was used as a secondary method after a primary ablative procedure had not been successful or if outcome data overlapped with subsequent larger series. In articles that described a combination of various techniques for selective reduction, cases were included only when BCO or RFA was used as the sole technique.
For each article, the number of cases, the median gestational age at procedure and delivery and the numbers of live births and neonatal deaths were collected. Overall survival was calculated by the exclusion of the neonatal deaths from the number of live births. The type of monochorionic pregnancy and the rates of cotwin death, miscarriage at <24 weeks of gestation, and preterm delivery at <28 and <32 weeks of gestation were also recorded. The indications for selective reduction, failed procedure rates and the reasons for this, maternal and fetal complications, rates of postprocedure brain injury, and neurodevelopmental outcomes were also analyzed.
Indications for selective reduction were subdivided into the primary indication for the procedure: twin-twin transfusion syndrome (TTTS), twin reversed arterial perfusion (TRAP), or other (included discordant structural anomalies, selective growth restriction, and multifetal pregnancy reduction). Overall survival, cotwin death, live birth, and neonatal death rates were stratified by indication wherever possible. Metaanalysis was performed using Review Manager (RevMan 2014, version 5.3; The Nordic Cochrane Centre, The Cochrane Collaboration, Copenhagen, Denmark). The Metaanalysis of Observational Studies in Epidemiology (MOOSE) statement was consulted during the study design.
For count data, Poisson regression was used to analyze outcomes for RFA compared with BCO (with the use of incidence risk ratios (IRRs) that were relative to BCO, which was regarded as the reference standard). Median gestational age at procedure and at delivery was extracted from each included article, was assessed for normality, and was reported as mean median (± standard deviation).
Results
Two hundred thirty-one articles initially were retrieved with the use of the previously detailed search strategy. All abstracts were screened, after which 28 full-text articles that were deemed to be relevant to this systematic review/metaanalysis were then reviewed. Of these, a further 8 articles were excluded because of overlap of subjects in subsequent publications from the same group. Three articles were excluded because it was not possible to define outcome data accurately from the pooled results that had been published in these 3 studies ( Figure 1 ).
After all relevant exclusions, 17 studies with 481 cases of BCO and 320 cases of RFA were available for analysis ( Table 1 ). Six articles reported outcomes that were specific for RFA, 8 articles reported BCO outcomes and 3 articles reported outcomes when both procedures were used in the same series. The number of cases, the median gestational age at procedure and delivery, the live birth rate, and the overall survival rate that were relevant to the individual studies are displayed in Table 1 .
| Study | Technique | Cases, n | Median gestational age, wk | Live birth, % | Overall survival, % | |
|---|---|---|---|---|---|---|
| At procedure | At delivery | |||||
| van Den Bos et al (2013) | RFA | 11 | 15.0 | 34.0 | 63.6 | 63.6 |
| BCO | 36 | 20.5 | 35.5 | 86.1 | 77.8 | |
| Bebbington et al (2012) | RFA | 58 | 20.2 | 33.0 | 82.8 | 70.7 |
| BCO | 88 | 20.9 | 34.7 | 94.3 | 85.2 | |
| Roman et al (2009) a | RFA | 20 | 20.3 | 36.0 | 95.8 | 87.5 |
| BCO | 40 | 21.5 | 39.0 | 92.8 | 88.0 | |
| Kumar et al (2014) | RFA | 100 | 17.96 | 35.2 | 78.0 | 78.0 |
| Berg et al (2014) | RFA | 7 | 23.0 | 32.3 | 100.0 | 85.7 |
| Lee et al (2013) | RFA | 98 | 20.2 | 37.0 | 83.0 | 80.0 |
| Lu et al (2013) a | RFA | 10 | 15.6 | 35.9 | 81.8 | NA |
| Cabassa et al (2012) | RFA | 7 | 17.4 | 33.0 | 71.0 | 71.0 |
| Moise et al (2008) | RFA | 9 | 19.5 | 36.1 | 66.0 | 66.0 |
| Nobili et al (2013) | BCO | 22 | 20.0 | 35.0 | 90.9 | NA |
| Lanna et al (2012) | BCO | 118 | 22.0 | 34.0 | 80.0 | 71.1 |
| Sundberg et al (2012) a | BCO | 65 | 20.4 | 35.4 | NA | 82.0 |
| He et al (2010) | BCO | 14 | 23.8 | 35.4 | 92.0 | 85.0 |
| Lewi et al (2013) a | BCO | 25 | 22.4 | NA | 92.0 | 88.0 |
| Young et al (2006) | BCO | 12 | 20.0 | 35.5 | 100 | 91.7 |
| Robyr et al (2005) | BCO | 46 | 20.0 | 34.0 | 82.6 | 74.0 |
| Taylor et al (2002) | BCO | 15 | 21.0 | 33.0 | 93.3 | 86.7 |
a Indicates studies in which triplet pregnancies are included and, if appropriate, in which 2 live births were used to calculate survival rates.
The mean median gestational age at procedure was 18.8 ± 2.5 weeks in the RFA group and 21.1 ± 1.2 weeks in the BCO group. This difference was found to be statistically significant ( P = .03). Although all cases that were reported pertained to monochorionic pregnancies with the majority being monochorionic twins, some studies included triplets (either monochorionic triplets or triplets with a monochorionic twin pair). Triplets accounted for 14.4% (46/320) of the RFA cohort and 4.6% (21/456) of the BCO cohort. It was not possible to stratify outcome by the type of multiple pregnancy for most of the included studies.
Pooled data for all the major outcomes are displayed in Table 2 . Because not all outcomes of interest were extractable from all 17 studies, the numbers of cases reflect those outcomes only when they were reported specifically in each article.
| Variable | Bipolar cord occlusion, n/N (%; 95% CI) | Radiofrequency ablation, n/N (%; 95% CI) | Incidence risk ratio (95% CI) | P value |
|---|---|---|---|---|
| Cotwin death | 51/481 (10.6; 8.0–13.9%) | 47/320 (14.7; 11.0–19.5%) | 1.38 (0.93–2.05) | .11 |
| Neonatal death b | 32/394 (8.1; 5.7–11.5%) | 14/310 (4.5; 2.7–7.6%) | 0.56 (0.30–1.04) | .07 |
| Live birth | 417/481 (86.7; 78.7–95.4%) | 260/320 (81.3; 71.9–91.7%) | 0.93 (0.80–1.09) | .41 |
| Overall survival c | 363/459 (79.1; 71.3–87.5%) | 238/310 (76.8; 67.6–87.2%) | 0.97 (0.82–1.14) | .72 |
| Preterm premature rupture of membranes d | 104/369 (28.2; 23.2–34.2%) | 39/220 (17.7; 12.9–24.3%) | 0.63 (0.43–0.91) | .01 |
| Premature delivery <28 wk e | 37/266 (13.9; 10.7–19.2%) | 29/211 (13.7; 9.5–19.8%) | 0.99 (0.61–1.61) | .96 |
| Premature delivery <32 wk f | 114/398 (28.6; 23.6–34.4%) | 50/211 (23.7; 17,9–31.3%) | 0.83 (0.59–1.15) | .26 |
| Miscarriage g | 8/255 (3.1; 1.5–6.3%) | 7/133 (5.3; 2.5–11.0%) | 1.67 (0.61–4.63) | .32 |
a Summary statistics for outcomes recorded from the 17 included articles. Proportions, incidence risk ratios and 95% confidence intervals, and probability values are derived from a comparison of outcomes of radiofrequency ablation (RFA) and bipolar cord occlusion (BCO) with the use of Poisson regression, with BCO as the reference category
b Data reported in 8 RFA articles and 9 BCO articles
c Data reported in 8 RFA articles and 10 BCO articles
d Data reported in 8 RFA articles and 8 BCO articles
e Data reported in 7 RFA articles and 6 BCO articles
f Data reported in 7 RFA articles and 8 BCO articles
Cotwin death and live birth rates were reported in all articles. The overall rate of cotwin death was 14.7% (11.0-19.5%; 47/320) in the RFA group and 10.6% (8.0-13.9%; 51/481) in the BCO cohort (IRR, 1.38; 95% confidence interval [CI], 0.93–2.05; P = .11). The live birth rate was 81.3% (71.9-91.7%; 260/320) for RFA cases and 86.7% (78.7-95.4%; 417/481) in the BCO group (IRR, 0.93; 95% CI, 0.80–1.09; P = .41).
The overall neonatal death rates in the RFA and BCO cohorts were 4.5% (2.7-7.6%; 14/310) and 8.1% (5.7-11.5%; 32/394), respectively (IRR, 0.56; 95% CI, 0.30–1.04; P = .07). The overall survival rate (live births excluding neonatal deaths) was 76.8% (67.6-87.2%; 238/310) in the RFA group and 79.1% (71.3-87.5%; 363/459) in the BCO cohort (IRR, 0.97; 95% CI, 0.82–1.14; P = .72). Data regarding miscarriage or delivery <24 weeks of gestation was available only for 5 of the studies in each group. Miscarriage occurred in 5.3% (2.5-11%; 7/133) of RFA cases and 3.1% (1.5-6.3%; 8/255) in the BCO group (IRR, 1.67; 95% CI, 0.61–4.63; P = .32).
Procedure-related complications were high in both groups, with a PPROM rate of 17.7% (12.9-24.3%; 39/220) in the RFA group and 28.2% (23.2-34.2%; 104/369) in the BCO group (IRR, 0.63; 95% CI, 0.43–0.91; P = .01). Rates of PPROM within 2 weeks of the procedure were reported in only 5 RFA and 4 BCO articles. Not all studies reported this specific interval between procedures to PPROM. Of 53 cases in the RFA cohort, only 2 experienced ruptured membranes within this time frame (3.8%; 95% CI, 0.94–15.1); of 184 cases in the BCO cohort, only 21 had ruptured membranes in the first 2 weeks (11.4%; 95% CI, 7.4–17.5; IRR, 0.33; 95% CI, 0.08–1.41).
The mean median gestational age at delivery was 34.7 ± 1.7 weeks in the RFA group and 35.1 ± 1.6 weeks in the BCO group. Pooled data for the overall rate of preterm delivery at <28 weeks of gestation was 13.9% (10.7-19.2%; 37/266) in the BCO group and 13.7% (9.5-19.8%; 29/211) in the RFA group (IRR, 0.99; 95% CI, 0.61–1.61; P = .96); for delivery at <32 weeks of gestation, data were 23.7% (17.9-31.3%; 50/211) in the RFA group and 28.6% (23.6-34.4%; 114/398) in the BCO cohort (IRR, 0.83; 95% CI, 0.59–1.15; P = .26).
The indications for selective reduction were ascertainable in only 15 of the 17 studies ( Table 3 ). The most frequent indication for RFA was TRAP in 48.8% of cases, compared with only 16.0% in the BCO cohort. Not all studies reported outcomes that were stratified by indication, so the survival data in Table 3 reflects this. Overall survival after selective reduction with the use of either procedure was lower when the indication was TTTS. It was not possible to stratify outcomes according to the severity of the TTTS because of the heterogeneous manner of reporting TTTS staging across the different studies.
| Procedure | Indication | n/N (%) | Overall survival, n/N (%) |
|---|---|---|---|
| Radiofrequency ablation | Twin-twin transfusion syndrome | 54/320 (16.9) | 33/49 (67.3) |
| Twin reversed arterial perfusion | 156/320 (48.8) | 121/153 (79.1) | |
| Other | 110/320 (34.4) | 72/94 (76.6) | |
| Bipolar cord occlusion | Twin-twin transfusion syndrome | 202/481 (42) | 149/194 (76.8) |
| Twin reversed arterial perfusion | 77/481 (16.0) | 58/73 (79.5) | |
| Other a | 155/481 (32.2) | 97/127 (76.4) | |
| Not reported b | 47/481 (9.8) |
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
