Learning curve for the perinatal outcomes of radiofrequency ablation for selective fetal reduction: a single-center, 10-year experience from 2013 to 2023

Introduction

Over the past few decades, assisted reproductive technologies have advanced greatly, leading to a substantial increase in the prevalence of multiple pregnancies. These pregnancies are associated with high risks of maternal and perinatal complications [1], 2] and thus are difficult to manage. Approximately 30 % of multiple pregnancies are monochorionic (MC) twin pregnancies [3], 4]. Furthermore, approximately one-third of MC twins, who share a placenta and have vascular anastomoses, may develop severe complications, including twin-to-twin transfusion syndrome (TTTS), selective fetal growth restriction (sFGR), twin reversed arterial perfusion sequence (TRAPS), twin anemia polycythemia sequence (TAPS), and discordant anomalies. These complications substantially increase the risks of perinatal morbidity and mortality, including preterm premature rupture of membranes (PPROM), intrauterine fetal death (IUFD), preterm birth, stillbirth, and neonatal death [1], 3], 5].

In MC twin pregnancies, due to the presence of placental vascular anastomoses, a life-threatening condition of one twin causes abrupt blood flow from the cotwin into the affected twin through the connecting vessels. This imbalance in circulation can trigger an immediate hypotensive and hypoxic event, resulting in the intrauterine death of or neurological injury to the unaffected twin [6], [7], [8]. Thus, selective reduction (SR) may be considered in severe fetal conditions to reduce the risk of harm, prevent the loss of the cotwin, and improve the outcomes of pregnancy [1], 9]. In addition, SR can be performed in cases of discordant fetal anomalies, high-order multiple gestations, maternal comorbidities, or even for social reasons to improve the prognosis of the surviving fetus or reduce the complications and risks associated with MC twin pregnancies.

SR can be performed using several techniques, such as radiofrequency ablation (RFA), bipolar cord coagulation, microwave ablation, and interstitial laser coagulation [1], [10], [11], [12]. RFA is associated with better perinatal outcomes because it is a simple and fast procedure and the instrument used in this procedure has a small diameter [9], 12], 13].

Although numerous studies have examined the outcomes of RFA for SR and identified the factors affecting these outcomes, the results remain debatable. This variability may be attributed to the differences in the study populations and sample sizes, inconsistent surgical indications, diverse clinical conditions, and varying technical expertise levels across different centers. Since the introduction of RFA for SR in our center in 2013, we have gained extensive experience and refined our techniques. Thus, we conducted this retrospective study to evaluate and compare the perinatal outcomes of SR using RFA in complex MC pregnancies and identify factors affecting these outcomes. In addition, we present our center’s learning curve to enhance the application of RFA and provide clinical insights and recommendations for patients.

Materials and methods

Results

Perinatal outcomes of RFA according to indications

As presented in Table 3, the GA at the time of the procedure for the EFR group (17.64 [17.04–18.96] weeks) was the smallest among the six groups (p=0.005). The survival rates in the TTTS, sFGR, TRAPS, fetal discordant anomaly, EFR, and TAPS groups were 75.0 % (30/40), 100.0 % (17/17), 75.0 % (9/12), 80.0 % (20/25), 100.0 % (10/10), and 100.0 % (3/3), respectively, with the TTTS and TRAPS groups having the lowest rates. The GA at delivery for the TTTS group was the smallest among the six groups (33.43 [27.57–37.21] weeks, p=0.021). The procedure-to-delivery interval was the shortest (60.0 [25.5–103.8] weeks) in the TTTS group and the longest (135.0 [126.5–143.3] weeks, p<0.001) in the EFR group among the six groups. Of the four cases of PPROM before 28 weeks, three occurred in the TTTS group, with two within one week of RFA, and one in the EFR group. Moreover, of the eight miscarriages, five were in the TTTS group and three in the TRAPS group. No significant differences were noted in PPROM before 28 weeks, PPROM within two weeks after RFA, IUFD, miscarriage, TOP, and stillbirth across all the six groups (p>0.05). The Kaplan–Meier curves demonstrated that the TTTS group had a significantly lower proportion of ongoing pregnancies and a higher proportion of early deliveries after RFA than the other groups (Figures 1 and 2).

Table 3:

Perinatal outcomes and post-procedure complications based on indications.

Variable	TTTS (n=40)	sFGR (n=17)	TRAPS (n=12)	Discordant anomalies (n=25)	EFR (n=10)	TAPS (n=3)	p-Value
GA at operation, weeks	22.43 (19.71–24.54)	20.71 (17.64–25.14)	21.29 (17.57–23.04)	22.43 (21.21–25.29)	17.64 (17.04–18.96)	24.57 (22.57–25.21)	0.005
GA at delivery, weeks	33.43 (27.57–37.21)	37.71 (34.64–38.93)	36.14 (28.75–37.82)	35.29 (31.43–36.64)	37.21 (35.93–38.14)	34.57 (33.79–35.43)	0.021
Pregnancies with live birth	30 (75.0 %)	17 (100.0 %)	9 (75.0 %)	20 (80.0 %)	10 (100.0 %)	3 (100.0)	0.097
GA at live delivery, weeks	34.81 ± 3.72	36.86 ± 2.54	37.02 ± 2.26	35.64 ± 2.94	36.91 ± 1.72	34.62 ± 1.64	0.125
GA at live birth ≤33+6 weeks	11 (36.7 %)	2 (11.8 %)	0 (0.0 %)	4 (20.0 %)	1 (10.0 %)	1 (33.3 %)	0.127
GA at live birth 34–36+6 weeks	7 (23.3 %)	4 (23.5 %)	3 (33.3 %)	10 (50.0 %)	3 (30.0 %)	2 (66.7 %)	0.280
GA at live birth ≥37 weeks	12 (40.0 %)	11 (64.7 %)	6 (66.7 %)	6 (30.0 %)	6 (60.0 %)	0 (0.0 %)	0.086
Operation-to-delivery interval, days	60.0 (25.5–103.8)	112.0 (77.0–137.5)	91.0 (51.0–139.0)	81.0 (48.5–109.5)	135.0 (126.5–143.3)	73.0 (71.5–80.0)	0.000
PPROM (<28 weeks)	3 (7.5 %)	0 (0.0 %)	0 (0.0 %)	0 (0.0 %)	1 (10 %)	0 (0.0 %)	0.414
PPROM within two weeks post-RFA	2 (5.0 %)	0 (0.0 %)	0 (0.0 %)	0 (0.0 %)	1 (10 %)	0 (0.0 %)	0.440
IUFD	2 (5.0 %)	0 (0.0 %)	0 (0.0 %)	3 (12.0 %)	0 (0.0 %)	0 (0.0 %)	0.578
IUFD within one week post-RFA	1 (2.5 %)	0 (0.0 %)	0 (0.0 %)	2 (8.0 %)	0 (0.0 %)	0 (0.0 %)	0.749
Miscarriage	5 (12.5 %)	0 (0.0 %)	3 (25.0 %)	0 (0.0 %)	0 (0.0 %)	0 (0.0 %)	0.063
TOP	2 (5.0 %)	0 (0.0 %)	0 (0.0 %)	2 (8.0 %)	0 (0.0 %)	0 (0.0 %)	0.888
Stillbirth (≥28 weeks)	1 (2.5 %)	0 (0.0 %)	0 (0.0 %)	0 (0.0 %)	0 (0.0 %)	0 (0.0 %)	1.000

1. Data are given as mean ± SD, n (%), or median (interquartile range). GA, gestational age; RFA, radiofrequency ablation; TTTS, twin-to-twin transfusion syndrome; sFGR, selective fetal growth restriction; TRAPS, twin reversed arterial perfusion sequence; EFR, elective fetal reduction; TAPS, twin anemia–polycythemia sequence; PPROM, preterm premature rupture of membranes; IUFD, intrauterine fetal death; TOP, termination of pregnancy. The definition for the significance of bold value is p<0.05.

Figure 1:

Kaplan–Meier survival curves of different indications based on the procedure-to-delivery interval. The TTTS group had a significantly lower proportion of remaining undelivered after RFA compared to other groups.

Figure 2:

The Kaplan–Meier survival curves of different indications based on the GA of delivery. The TTTS group had a significantly higher proportion delivered at early GA after RFA compared to other groups.

Comparison of outcomes throughout various time periods and learning curve

As depicted in Figure 3, the locally weighted scatterplot smoothing illustrates changes in the cotwin loss rate and caseload over the study period. During the 10-year period, the overall cotwin loss rate decreased to 16.8 %, with a notable decline in the earlier years before stabilizing around 2018. In addition, the annual survival rate fluctuated but gradually increased, reaching approximately 90 % in the last three years (Figure 4). Accordingly, we divided the cases into two periods for analysis: period 1 (2013–2018) and period 2 (2019–2023). Table 5 presents a comparison of perinatal outcomes between the two periods.

Figure 3:

The changes of cotwin loss rate and caseload over the past years.

Figure 4:

Survival rate in each year during the study period.

Table 5:

Comparison of obstetric characteristics and perinatal outcomes following RFA of the two periods.

Variable	Period 1 (n=47)	Period 2 (n=60)	p-Value
Maternal age at operation, years	29.00 (27.00–33.00)	30.00 (28.00–32.00)	0.275
BMI, kg/m2	24.54 (22.12–26.67)	24.17 (22.76–27.17)	0.853
Gravidity, time	2.0 (1.0–2.0)	2.0 (1.0–2.0)	0.81
Parity, time	0.0 (0.0–1.0)	0.0 (0.0–1.0)	0.937
GA at operation, weeks	22.86 (19.00–25.71)	21.36 (18.07–23.54)	0.032
ART	1 (2.1 %)	10 (16.7 %)	0.021
Chorioamnionicity
MCDA	44 (93.6 %)	47 (78.3 %)	0.058
MCMA	1 (2.1 %)	1 (1.7 %)
MCTA	2 (4.3 %)	7 (11.7 %)
DCTA	0.0 (0.0–1.0)	5 (8.3 %)
Type of multiple pregnancies
Twins	45 (95.7 %)	48 (80.0 %)	0.020
Triplets	2 (4.3 %)	12 (20.0 %)
Indication for RFA, n (%)
TTTS	19 (40.4 %)	21 (35.0 %)	0.173
sFGR	6 (12.8 %)	11 (18.3 %)
TRAPS	8 (17.0 %)	4 (6.7 %)
Discordant anomalies	12 (25.5 %)	13 (21.7 %)
EFR	2 (4.3 %)	8 (13.3 %)
TAPS	0 (0.0 %)	3 (5.0 %)
Perinatal outcome
GA at delivery, weeks	34.71 (28.00–37.29)	36.00 (33.75–37.86)	0.133
Interval from operation to delivery, days	73.00 (23.00–101.00)	100.00 (70.00–129.00)	0.003
Pregnancies with survival	34 (72.3 %)	55 (91.7 %)	0.010
GA at live delivery, weeks	37.00 (34.39–38.04)	36.29 (34.00–37.86)	0.836
GA at live birth ≤33+6 weeks	8 (23.5 %)	11 (20.0 %)	0.792
GA at live birth 34–36+6 weeks	8 (23.5 %)	21 (38.2 %)	0.171
GA at live birth ≥37 weeks	18 (52.9 %)	23 (41.8 %)	0.383
Interval from operation to live delivery, days	90.41 ± 38.40	102.00 ± 35.08	0.148
PPROM (<28 weeks)	3 (6.4 %)	1 (1.7 %)	0.318
PPROM within two weeks post-RFA	3 (6.4 %)	0 (0.0 %)	0.082
IUFD	3 (6.4 %)	2 (3.3 %)	0.652
IUFD within one week post-RFA	3 (6.4 %)	0 (0.0 %)	0.082
Miscarriage	8 (17.0 %)	0 (0.0 %)	0.001
TOP	2 (4.3 %)	2 (3.3 %)	1.000
Stillbirth (≥28 weeks)	1 (2.1 %)	0 (0.0 %)	NA

1. Data are given as median (interquartile range), or n (%). GA, gestational age; BMI, body mass index; ART, assisted reproduction techniques; MCDA, monochorionic diamniotic; DCDA, dichorionic diamniotic; MCMA, monochorionic monoamniotic; MCTA, monochorionic triamniotic; DCTA, dichorionic triamniotic; TTTS, twin-to-twin transfusion syndrome; sFGR, selective fetal growth restriction; TRAPS, twin reversed arterial perfusion sequence; EFR, elective fetal reduction; TAPS, twin anemia–polycythemia sequence; PPROM, preterm premature rupture of membranes; IUFD, intrauterine fetal death; TOP, termination of pregnancy. The definition for the significance of bold value is p<0.05.

The survival rate in period 2 was significantly higher than that in period 1 (72.3 vs. 91.7 %, p=0.01), with the procedure-to-delivery interval also being higher in period 2 [73.00 (23.00–101.00) days vs. 100.00 (70.00–129.00) days, p=0.003]. The miscarriage rate in period 2 declined to 0.0 % from 17.0 % in period 1 (p=0.001), with the only stillbirth occurring in period 1. Rates of PPROM before 28 weeks, PPROM within two weeks of RFA, cotwin IUFD, and TOP did not significantly differ between the groups (p>0.05). The Kaplan–Meier curves indicated a higher proportion of ongoing pregnancies following RFA in period 2 than in period 1 (p=0.021, Figure 5).

Figure 5:

Kaplan–Meier survival curves of two periods after RAF. There was a higher proportion of undelivered pregnancies following RFA in period 2 compared with period 1 (p=0.021).

Table 6 presents a comparison of the perinatal outcomes between the two periods (2013–2018 vs. 2019–2023) for each indication, showing significant differences only in TTTS as the indication of RFA. In the TTTS group, the procedure-to-delivery interval was significantly longer in period 2 than in period 1 (29.00 [21.00–90.00] days vs. 95.00 [42.00–121.50] days, p=0.006), and the survival rate increased from 63.2 % in period 1–85.7 % in period 2. In addition, we noted a decrease in the miscarriage rate in period 2 (26.3 vs. 0.0 %, p=0.018) compared with period 1.

Table 6:

Comparison of the perinatal outcomes of two periods according to indications.

Perinatal outcomes	Period	TTTS	%	SIUGR	%	TRAPS	%	Discordant anomalies	%	EFR	%	TAPS	%
Live birth (n=89)	Period 1	12/19	63.2 %	6/6	100.0 %	5/8	62.5 %	9/12	75.0 %	2/2	100.0 %	0/0	–
	Period 2	18/21	85.7 %	11/11	100.0 %	4/4	100.0 %	11/13	84.6 %	8/8	100.0 %	3/3	100.0 %
	p-value	0.148		–		0.491		0.645		–		–
GA at delivery, weeks	Period 1	31.00 (26.14–35.57)		38.50 (37.54–39.00)		37.00 (24.97–39.18)		34.93 (29.79–36.75)		35.43 (33.85–37.00)		–
	Period 2	34.28 (30.00–37.93)		36.57 (34.00–37.86)		34.79 (34.07–36.79)		35.85 (32.93–37.64)		37.64 (36.43–38.14)		34.57 (33.79–35.43)
	p-value	0.083		0.098		0.808		0.470		0.267		–
Interval from operation to delivery, days	Period 1	29.00 (21.00–90.00)		97.00 (78.25–138.75)		91.00 (30.00–141.25)		72.00 (9.50–106.75)		126.50 (116.00–137.00)		–
	Period 2	95.00 (42.00–121.50)		120.00 (77.00–140.00)		98.00 (69.25–136.50)		83.00 (61.00–113.00)		135.50 (130.50–147.75)		73.00 (71.50–80.00)
	p-value	0.006		0.884		1.000		0.270		0.533		–
PPROM within two weeks post-RFA (n=3)	Period 1	2/19	10.5 %	–	–	–	–	–	–	1/2	50.0 %	–	–
	Period 2	0/21	0.0 %	–	–	–	–	–	–	0/8	0.0 %	–	–
	p-value	0.219		–		–		–		0.200		–
IUFD with one week (n=3)	Period 1	1/19	5.3 %	–	–	–	–	2/12	16.7 %	–	–	–	–
	Period 2	0/21	0.0 %	–	–	–	–	0/13	0.0 %	–	–	–	–
	p-value	0.475		–		–		0.220		–		–
Miscarriage (n=8)	Period 1	5/19	26.3 %	–	–	3/8	37.5 %	–	–	–	–	–	–
	Period 2	0/21	0.0 %	–	–	0/4	0.0 %	–	–	–	–	–	–
	p-value	0.018		–		0.491		–		–		–
TOP (n=4)	Period 1	1/19	5.3 %	–	–	–	–	1/12	8.3 %	–	–	–	–
	Period 2	1/21	4.8 %	–	–	–	–	1/13	7.7 %	–	–	–	–
	p-value	1.000		–		–		1.000		–		–
Stillbirth (≥28 weeks) (n=1)	Period 1	1/19	5.3 %	–	–	–	–	–	–	–	–	–	–
	Period 2	0/21	–	–	–	–	–	–	–	–	–	–	–
	p-value	0.475		–		–		–		–		–

1. Data are given as median (interquartile range), or n (%). GA, gestational age; RFA, radiofrequency ablation; MCDA, monochorionic diamniotic; DCDA, dichorionic diamniotic; MCMA, monochorionic monoamniotic; MCTA, monochorionic triamniotic; DCTA, dichorionic triamniotic; TTTS, twin-to-twin transfusion syndrome; sFGR, selective fetal growth restriction; TRAPS, twin reversed arterial perfusion sequence; EFR, elective fetal reduction; TAPS, twin anemia–polycythemia sequence; PPROM, preterm premature rupture of membranes; IUFD, intrauterine fetal death; TOP, termination of pregnancy. The definition for the significance of bold value is p<0.05.

Discussion

RFA, a minimally invasive technique characterized by ease of access and application, has become a well-established method for managing complex MC pregnancies.

In their systematic review and meta-analysis, Gaerty et al. [11] reported a survival rate of 76.8 % (67.6–87.2 %, 238/310) for RFA reduction in MC pregnancies. Similarly, Li et al. [14] reviewed the studies on RFA reduction in multiple pregnancies conducted from 2008 to 2021 and determined the mean and maximum survival rates of 76.55 % ± 8.01 % and 91.9 % [13], respectively. They also reported their survival rate as 78.76 %. The overall survival rate in the current study was 82.3 %, which is comparable to that reported in previous studies. These results collectively indicate that RFA is an effective approach for SR in complicated MC multiple pregnancies, consistently demonstrating a favorable cotwin survival rate of over 70 %.

In our study, the median GA at delivery was 35.4 weeks, which is in agreement with the findings of Gaerty et al. [11], who reported a mean GA at delivery of 34.7 ± 1.7 weeks, and Li et al. [14], who reported the median GA at delivery in the range of 31.6–36.8 weeks.

PPROM is a major concern following intrauterine inventions. In this study, three of the four cases of PPROM before 28 weeks occurred within two weeks of RFA, potentially contributing to miscarriages and other adverse outcomes. In addition, IUFD is an undesirable outcome of RFA, typically occurring within 24–48 h of surgery; however, it can also occur weeks later [15], 16]. Previous studies have indicated that the IUFD rate ranges from 4.5 to 14.7 % [11], 14], [17], [18], [19], and in this study, this rate was 4.7 % (5/107), with two cases occurring within 48 h following RFA. Early IUFD likely occurs due to interfetal hemorrhage and incomplete vessel occlusion, leading to transfusion of the cotwin’s blood to the moribund fetus [12], 15], 20]. The causes of later IUFD instances remain unclear. Moreover, miscarriage, with reported incidence rates ranging from 5.3 to 17.3 % [11], [21], [22], [23], is another undesirable outcome of RFA. Some studies have suggested that the location of the targeted fetus and the indications for RFA can affect the likelihood of miscarriage [21], 24]. In our cohort, all eight miscarriage cases were associated with TTTS or TRAPS as indications for RFA, aligning with the findings of previous studies. However, additional studies are warranted to clarify these relationships.

In the current cohort, the most common indication for RFA was TTTS, followed by discordant anomalies, sFGR, and TRAPS. Li et al. [14] demonstrated that cotwin loss after RFA may be associated with the indications for RFA. In addition, our findings indicated that indications can affect the perinatal outcomes of the procedure.

In our study, the TTTS and TRAPS groups had the lowest overall survival rate (75.0 %). Furthermore, among the six groups, the TTTS group exhibited the worst perinatal outcomes, with the smallest GA at delivery, the shortest RFA to delivery interval, and a higher frequency of adverse consequences. A meta-analysis by Wang et al. [25] reported a survival rate of 69.3 % for TTTS, which is significantly lower than that for other RFA indications. Similarly, a systematic review by Donepudi et al. [26] reported a lower cotwin survival rate for TTTS cases undergoing RFA. Previous systematic reviews by Gaerty et al. [11] and others [22], 23], 27], 28] have also indicated that TTTS outcomes were less favorable than those of other RFA indications. This trend is likely related to the pathophysiological mechanisms of TTTS, where the hemodynamic imbalance may adversely affect both fetuses. The donor twin experiences hypovolemia, oliguria, and oligohydramnios, potentially leading to hypertension and renal dysfunction, whereas the recipient twin experiences hypervolemia and circulatory overload, which can cause polyuria and polyhydramnios, ultimately leading to hypertension and cardiac dysfunction [3], 29], 30]. Thus, the morbidity of the cotwin might not be preventable through the reduction of the affected twin. In addition, the high amniotic fluid pressure caused by severe polyhydramnios increases the risks of PPROM, miscarriage, and preterm delivery [31].

Fetoscopic laser photocoagulation is the preferred treatment for TTTS [8], and RFA is frequently preferred for cases where laser therapy is not feasible, due to factors such as the advanced Quintero stage, proximity of umbilical cord insertion sites, poorly accessible anastomoses owing to the position of the placenta or fetuses, and parental preference. At our facility, we provide individualized counseling to all pregnant women with TTTS and their families based on the specific clinical situation of the woman and the fetuses. The procedure details and possible risks associated with both fetoscopic laser surgery and RFA are fully explained to the pregnant women and their families, as well as the possible outcomes of the fetuses after the procedures, and the choice is made by the women and their families after consideration. This study included 10 TTTS II cases, 19 TTTS III cases, and 11 TTTS IV cases, who preferred to undergo RFA owing to multiple factors, including fetal factors such as advanced stage, fetal anomaly, growth restriction, severe fetal hydrops and heart failure, maternal factors such as history of cesarean section, advanced age, and severe maternal complication, or concerns regarding the difficulties and risks of fetoscopic laser surgery, fetal prognosis, and costs (Supplementary Table 1).

Zhang et al. [32] reported a survival rate of 70 % for the pump twin in TRAPS cases treated with RFA, a finding echoed by Rahimi-Sharbaf et al. [33]. We obtained a similar result in our study. Treating TRAPS can be challenging due to significant tissue edema, potentially necessitating multiple circulations and a longer duration of operation [24], 34]. In addition, a pump twin experiencing persistently high output can lead to polyhydramnios, which can increase the risks of PPROM and miscarriage. In our study cohort, polyhydramnios was noted in two miscarriage cases in the TRAPS group, potentially contributing to adverse outcomes.

We identified factors affecting perinatal outcomes following RFA. Univariate logistic regression indicated that a cervical length of <25 mm after RFA may be the only risk factor for cotwin loss after RFA. Given the established association between short cervical length and increased risks of miscarriage and preterm birth [35], [36], [37], [38], cervical length may be a risk factor for unfavorable outcomes following RFA. However, in multiple pregnancies, its significance may be intertwined with other variables, and thus it may become nonsignificant after multivariate adjustment.

We also examined other potential factors affecting the perinatal outcomes, including various maternal and procedural characteristics. However, univariate logistic regression analysis revealed that the impact of these covariates was not discernible. The plausible explanation for the cotwin loss following RFA is the presence of several, interconnected, and mutually reinforcing factors, warranting additional studies to clarify the associations between various risk factors and the adverse outcomes after reduction by RFA.

Although the precise causes of cotwin loss after RFA remain to be clarified, we acknowledge the significant effects of proficiency in the use of refined techniques on perinatal outcomes following RFA. Thus, we reviewed data from our center, spanning 10 years since the introduction of RFA. During the study period, as case numbers increased, the annual survival rate was found to gradually increase. Comparing the perinatal outcomes from the earlier years with the recent ones, we noted a significant increase in survival rates, an extension in the procedure-to-delivery interval, and a reduction in miscarriage rates across the cohort. A sub-analysis based on indications revealed a similar increase in the procedure-to-delivery interval and a decrease in miscarriage rates, particularly in TTTS cases. This significant increase in the survival rate may be partially attributed to advancements in antenatal management and ultrasound technology for early diagnosis, as well as to a learning curve effect in the RFA procedure. The enhanced perinatal outcomes, especially in TTTS cases, provide strong evidence of our clinical progress.

At the inception of the application of a new technique, the accumulation of cases and procedural experiences contributes to the enhancement of technical proficiency, as observed in the first few years of period 1, with further improvements in perinatal outcomes, as observed in period 2. The dynamic improvement in survival rates could be due to the initial tacking of technically simpler cases, then progressing to more challenging ones, which might temporarily lower survival rates until technical improvements lead to subsequent increases. The findings demonstrated that as a center’s experience enriches, the survival rate increases significantly, indicating the effectiveness of refined techniques in improving perinatal outcomes following RFA. These results emphasize the significance of mastering advanced techniques in optimizing RFA treatment outcomes.

The strengths of our study include the inclusion of consecutive cases over a decade, providing a relatively large sample size. In addition, we covered nearly all types of MC pregnancy complications, enabling comparisons of perinatal outcomes across various indications and an exploration of factors affecting RFA outcomes. Moreover, we demonstrated the expertise and learning curve of our center, which may be beneficial for other centers and serve as a reference to improve clinical management and patient counseling.

This study has some limitations that should be addressed. First, due to the retrospective nature of the study, the documentation of certain procedural details such as whether the instruments crossed the placenta during the operation was missing in some of the early cases. Besides, it has a observational design and was conducted in a single center. Moreover, because most of our patients were referred to our center from across the country and continued their perinatal care and delivery with their local obstetricians, the detailed information about delivery and newborns could not be obtained, and the long-term neurological development outcomes of the surviving fetus remain unknown. Additional large-scale multicenter prospective studies of thorough recording operative details during operations with the long-term follow-up of survivors’ neurodevelopmental outcomes should be conducted to identify factors affecting outcomes following RFA more comprehensively.

Conclusions

This study describes our 10-year experience of SR in complex MC pregnancies using RFA at a tertiary regional referral center for prenatal diagnosis and fetal medicine. Our findings highlight that RFA for fetal reduction is a safe and effective method for managing complicated MC pregnancies. The postoperative perinatal outcomes may be affected by the indications for RFA, with TTTS resulting in the most undesirable outcomes. The accumulation of experience with the increasing number of cases, contributing to improved technical proficiency, can further enhance the postprocedural outcomes of RFA.