Twin Reversed Arterial Perfusion Sequence

TRAP sequence is diagnosed by ultrasound demonstrating identical twins with one twin lacking a heart and multiple anomalies combined with an artery-to-artery connection between the twins. Doppler ultrasound detects a reversal of blood flow in the umbilical artery of the baby without a heart.

While some centers have considered the presence of TRAP sequence an indication for fetal intervention, it is not necessary in all cases. Because all fetal interventions have some risk, we reserve fetal intervention for pregnancies complicated by TRAP sequence at the greatest risk for complications. The best indicator of this risk is the size of the acardiac twin. The larger the acardiac twin, the greater the strain on the pump twin. In many cases of TRAP sequence, however, the acardiac twin remains small and does not cause a strain on the pump twin. We use of ratio of the estimated fetal weight of the acardiac twin to the estimated fetal weight of the pump twin to identify pregnancies at risk. Once the ratio of the acardiac-to-pump twin weight exceeds 0.5, it identifies a pregnancy with a 50% chance of adverse outcome indicating the need for closer monitoring. Once the ratio of the acardiac-to-pump twin weight exceeds 0.7, it identifies a pregnancy with a 90% chance of adverse pregnancy outcome without fetal intervention.

We also use the presence of polyhydramnios (too much amniotic fluid) and elevated fetal cardiac output on fetal echocardiography. However, these findings rarely occur in cases with acardiac-to-pump twin ratios < 0.7. Many cases never reach the 0.7 threshold, and we have followed pregnancies like this without intervention with uniform survival and an average gestational age at delivery of 38 weeks.

There have been a number of fetal interventions described to treat TRAP sequence, but it is clear that the best outcomes are achieved with ultrasound-guided intra-fetal radiofrequency ablation. Radiofrequency ablation (RFA) uses heat generated by radiowaves to coagulate, or clot off, the umbilical vessels of the acardiac twin at the umbilical cord insertion. Under ultrasound guidance, a 19-gauge needle is placed through the mother’s abdomen, through the uterine wall, into the amniotic sac, and into the abdominal wall of the acardiac twin at the umbilical cord insertion. The RFA device has multiple prongs or tines that are deployed around the umbilical cord vessels. The RFA is turned on for 2 minutes generating heat only around the tines, causing coagulation of the vessels. Usually after one application, blood flow can no longer be detected.

The first reported series of pump twin survival was noted to be 91%. In a larger multi-institution series, pump twin survival was reported to be 80%. This series included data from multiple institutions with variable experience and single vs. multi-pronged radiofrequency needles used. In Dr. Crombleholme’s experience with 100 cases of TRAP sequence, the pump twin survival has been 98%.

Van Allen et al., recommended a twin reversed arterial perfusion sequence to describe all acardiac fetuses. The TRAP sequence denotes a common pathophysiology for all forms and leads to an explanation of how a gradation of abnormalities can be produced (Van Allen et al. 1983). The fundamental requirement for the TRAP sequence is the development of arterial-to-arterial vascular anastomoses between the umbilical arteries of twins early in embryogenesis. The importance of vascular arterial anastomosis as the pathophysiology for acardia was first elucidated in 1879 by Ahlfeld (Ahlfeld 1879). The embryo with the hemodynamic advantage becomes the pump twin. The pump twin retrogradely perfuses the other twin with deoxygenated blood along the umbilical artery/arteries to the iliac artery/arteries to the abdominal aorta. The lower limbs and abdominal organs supplied by the iliac arteries and abdominal aorta preferentially receive a better blood supply, albeit with deoxygenated blood, and usually develop better than the upper part of the body (Van Allen et al. 1983). TRAP occurs in monochorionic pregnancies, and previous theories of polar body fertilization to explain acardius in sex-discordant twins have been discounted (Bieber et al. 1981; Fisk et al. 1996). In addition, the TRAP sequence is more common in monozygotic triplets than in monozygotic twins (James 1977; Healey 1994). Some authors have suggested a slight female preponderance in acardiac twins while other studies have not supported this idea (Healy 1994, James 1977).

The members of a TRAP sequence are known as the “perfused” twin and the “pump” twin. The perfused twin in a TRAP sequence is an example of the impact of vascular disruption on morphogenesis. Multisystem malformations, as well as unusual body form, are found in the perfused twin. The malformations found in cases of acardia include growth abnormalities, partial or complete absence of the cranial vault, anencephaly, holoprosencephaly, absent or rudimentary facial features, absent or rudimentary upper and/or lower limbs, absent lungs and heart, gastrointestinal atresia, omphalocele, gastroschisis, and absent liver, pancreas, spleen, and kidneys (Van Allen et al. 1983). Of 33 acardiac fetuses with a known karyotype, 11 (33%) were abnormal (Healey 1994). The karyotypic abnormalities included monosomy, trisomy, deletions, mosaicism, and polyploidy. The pattern of structural abnormalities found in the perfused twins with abnormal karyotypes is not appreciably different from those with normal karyotypes (Van Allen et al. 1983). Van Allen et al. suggested that the abnormal karyotype is not responsible for the malformation complex, but rather that it contributes to the discordant development between twins, increasing the likelihood of reversal of arterial blood flow if an anastomosis is present.

The presence of an acardiac twin requires a “pump” twin to provide circulation for itself as well as the acardiac co-twin. In many cases, the acardiac twin is almost equal in size to the normal twin. The pump twin is usually morphologically and genetically normal. In a review of 34 pump fetuses with a known karyotype, only 3 (8.8%) were abnormal as a result of trisomy (Healy 1994). The pump twin may show evidence of the physiologic consequence of fetal cardiac overload and congestive heart failure with hepatosplenomegaly. The principal perinatal problems associated with acardiac twinning are pump twin congestive heart failure, polyhydramnios, and preterm delivery (Moore et al. 1990).

The reported fetal/neonatal mortality in untreated TRAP sequence in the pump twin is substantial ranging from 50-75% (Gillim and Hendricks 1953; Napolitani and Schreiber 1960; Van Allen et al. 1984; Moore et al. 1990; Sogaard 1999). One factor thought to be contributing to the high perinatal mortality rate is the increased cardiac demands placed on the pump twin to perfuse the acardiac twin (Sullivan et al. 2003).

Premature delivery is another important factor determining the prognosis for the pump twin (Healey 1994; Moore et al. 1990; Van Allen et al. 1983). In one study where approximately 55% of acardiac pregnancies resulted in fetal or neonatal death, approximately one-quarter of the pregnancies were delivered after 36 weeks. Preterm delivery and the attendant long-term morbidities complicated the remaining quarter (Moore et al. 1990).

The incidence of the TRAP sequence is estimated as 1% of monozygotic twins, with birth estimates ranging from 1 in 35,000 to 1 in 50,000 births (D’Alton and Simpson 1995, Gillim and Hendricks 1953, Napolitani and Schreiber 1960). Acardia was observed in 1 of 606 twin pregnancies, and the rate of twins was calculated at 1 in 86.5 births in the United States (Gillim and Hendricks 1953). Van Allen et al. (1983) have suggested these figures to be a gross underestimate of the true frequency of the TRAP sequence because many cases may go unrecognized due to early pregnancy loss. In contrast, an analysis of data from the Eurocat Network (European Registration of Congenital Anomalies and Twins) gave a prevalence of acardia of 0.064 in 10,000 births, which is much lower than were previous estimates in the literature (Haring et al. 1993).

Ultrasonographic features useful in the diagnosis of acardia include an absence of normal cardiac structure and cardiac movement and variable structural abnormalities. Common structural abnormalities identified in the acardiac fetus include anencephaly, omphalocele, and absence of upper limbs. Most cases have edematous soft tissue, and large cystic hygroma-like spaces are commonly identified in the skin (Mack et al. 1982).

The placentation is most commonly monochorionic diamniotic (74%), in which a thin membrane will be seen dividing the sac of the acardiac fetus from the pump fetus (Healey 1994). Monoamnionicity is present in approximately 24% of cases (Healey 1994). In exceptional cases, dichorionicity may be diagnosed (Healey 1994). Polyhydramnios is common as are abnormalities in the umbilical cord or in its insertion (Dashe et al. 2001). The umbilical cord will demonstrate a single umbilical artery in approximately two-thirds of cases, and in one-third the number of cord vessels will be normal (Healey 1994). A velamentous insertion of the cord or a conjoined cord insertion may be present (Dashe et al. 2001).

Measurement of the acardiac twin should be performed, because the ratio of the weight of the acardiac twin to that of the pump twin is useful to predict pregnancy outcome. Because of the structural abnormalities, the biometric parameters of biparietal diameter, abdominal circumference, and femur length may not be available or reliable in an acardiac fetus. This problem of the antenatal determination of the acardiac twin’s weight has been addressed by Moore et al. (1990). The dimensions and weights of 23 acardiac twins were used for the analysis. A second-order regression equation (weight [g] = –1.66 × length + 1.21 × length2) was computed and was predictive of acardiac weight with the use of its longest linear measurement (r = .79; P < 0.001; SEE = 326 g). When the actual and equation-predicted weights were compared, the mean error (±SE) in prediction was 240 ± 156 g. Careful Doppler examination of the acardiac fetus will also demonstrate a reversal of flow in the umbilical artery of the acardiac fetus, with flow going from the placenta toward the acardiac fetus (Benson et al. 1989; Malone and D’Alton 2000).

The pump twin should have a detailed structural survey performed because trisomy has been reported in up to 9% of cases (Healey 1994), and sonographic features typical of a trisomic fetus may be identified. Fetal echocardiography is helpful in detecting early signs of in-utero congestive heart failure in the pump twin. Atrial and ventricular enlargement can be an initial feature of impending cardiac decompensation and can be measured using M-mode by obtaining a transverse view through the cardiac chambers (Allan 1986; DeVore 1987). The ventricular fractional shortening capacity can also be calculated using M-mode with the formula (D – S)D × 100, where D is the diastolic and S is the systolic ventricular size. A low value is indicative of poor cardiac contractility. A pericardial effusion may be present and is a sign of congestive heart failure. Tricuspid regurgitation, demonstrated by Doppler studies of the tricuspid valve, is also a sign of congestive heart failure (Shenker et al. 1988; Silverman et al. 1985). Combined ventricular output (CVO) can be measured to determine if the pump twin is in a high output state. In TRAP sequence at a gestational age too early to determine CVO, Kinsel-Ziter et al. have demonstrated a good correlation of increased CVO with increased cardiothoracic ratio allowing extrapolation to early gestation cases (Kinsel-Ziter et al. 2009).

Doppler studies should be performed in both the acardiac and pump twins. Verification of circulatory reversal by pulsed Doppler sonography of the acardiac twin can be documented with reversed direction of flow in the umbilical artery and vein (Benson et al. 1989; Dashe et al. 2001; Donnenfeld et al. 1991; Langlotz et al. 1991; Pretorius et al. 1988; Sherer et al. 1989). In the study by Dashe et al., between 1990 and 1997, Doppler studies were performed in 6 monochorionic pregnancies complicated by the TRAP sequence. Pulsatile flow in the umbilical vessels of the acardiac and pump twins were insonated. Reversal of flow in the acardiac twin was demonstrated in all cases. Resistive index values were calculated, and the difference in resistive index between the pump and acardiac twins was evaluated. In the acardiac twins, no ratio of systolic to diastolic velocity or resistive index value was associated with a good or with a poor prognosis for the pump twin. In the pump twins, resistive index differences > 0.20 between the pump and the acardiac twins were associated with good outcomes, while resistive index differences < 0.05 were associated with poor outcomes (Dashe et al. 2001).

Even though ultrasound remains the primary evaluation of density complicated by TRAP sequence, MRI does provide additional information that may be relevant to patient counseling as well as prognosis (Guimaraes 2011). In a series of 35 pregnancies complicated by TRAP sequence, Guimaraes et al., found 43% of pump twins showed signs of cardiac decompensation that correlated with fetal echocardiographic findings 93% of the time. The pump twin develops cardiac overload and TRAP sequence as well as chronic hypoxia caused by the mixing of returning venous blood from the acardiac fetus (Steffensen 2008).

In the series by Guimaraes et al., 1 fetus (3%) showed evidence of brain ischemia in the pump twin at presentation. The findings of fetal MRI largely confirmed findings on ultrasound, however, the field-of-view provided by fetal MRI facilitates operative planning for ultrasound-guided intra-fetal radiofrequency radiofrequency ablation procedures. The umbilical cord insertion and the acardiac twin are abnormal in 54% of cases, which further lends MRIs utility in operative planning.

In untreated TRAP sequence, the mortality ranges from 50 to 75% and is usually due to complications of prematurity, or high-output cardiac failure (Moore 1990, Healy 1994, Steffenson2008). Sonographic assessment of disease severity includes the ratio of the estimated fetal weight of the acardius-to-pump twin, the presence of polyhydramnios, umbilical artery resistance, and pulsatility index ratios. However, the earliest and most important signs of cardiovascular decompensation in the pump twin are best appreciated on fetal echocardiography.

In TRAP sequence, the pump twin must provide cardiac output to both its own systemic and placental circulations as well as the vascular circulation of the acardius. In addition, the venous return of the acardius is added to the pump twin circulation. Kinsel-Ziter et al., demonstrated a direct relationship between combined cardiac index (CCI) and cardiothoracic area and cardiac morphometric measurements. Elevated CCI was found to be associated with a high incidence of cardiovascular compromise as assessed by the cardiovascular profile score (CVPS) due to the presence of ventricular systolic dysfunction, atrioventricular valve incompetence, cardiomegaly, hydrops or abnormal Doppler velocimetry. An elevation of the CCI was found to be a more sensitive indicator of impending heart failure than an elevated acardius-to-pump twin weight ratio. This group also reported the echocardiographic response to the treatment of TRAP sequence by ultrasound-guided intra-fetal RFA demonstrating an acute volume unloading of the pump twin heart with normalization of the CCI postoperatively. Post-RFA fetal echocardiography demonstrated the ventricular systolic function rapidly returns to normal.

The acardiac fetus may be mistaken for an anencephalic fetus. The sonographic features of the absent trunk region in addition to increased soft tissue in the body aid in the correct diagnosis (Billah et al. 1984).

The TRAP sequence has been mistaken for intrauterine fetal demise (IUFD) of one twin in a multiple gestation (Malone and D’Alton 2000). Evidence of growth in the “dead” fetus and a “twitching” noted on repeat ultrasound examination has allowed the diagnosis of an acardiac twin to be made (Cardwell 1988). In a severely macerated fetus, the skeletal and visceral forms are more differentiated, and the soft-tissue edema is less advanced than in the case of acardia (Mack et al. 1982). The use of color-flow Doppler can assist in differentiating between a single IUFD in a co-twin and the TRAP sequence (Malone and D’Alton 2000). Pulsed Doppler examination has been used to demonstrate reversed flow through the umbilical artery of the acardiac twin (Pretorius et al. 1988).

The principal perinatal problems associated with acardiac twinning are pump twin congestive heart failure, polyhydramnios, and preterm delivery. The antenatal diagnosis of TRAP can be made through sonographic examination and has been reported in the literature only since 1980. A large series of acardiac twins have attempted to identify factors prognostic of favorable outcome for the pump twin (Healey 1994; Moore et al. 1990).

In the series of 49 cases reported by Moore et al. (1990), one-third of fetuses were delivered before they were viable. In this study, viability was defined as delivery at or beyond 25 weeks of gestation. Of the potentially viable 33 cases, 4 (12%) ended in the death of the pump twin in utero. The overall perinatal mortality was 55% and was primarily associated with prematurity.

Polyhydramnios is a common complication, occurring in 46% of all acardiac pregnancies. It is strongly associated with preterm labor and congestive heart failure in the pump twin. Eighty-two percent of patients with polyhydramnios experienced preterm labor requiring hospital admission and treatment, as compared with 22% of pregnancies with normal amniotic fluid (P < 0.01). Polyhydramnios was observed in 78% of pump twins with congestive heart failure as compared with 13% of those in whom congestive heart failure was not confirmed (P < 0.001). The perinatal outcome was strongly related to the ratio of the weight of the acardiac twin to that of the pump twin. The mean overall ratio of the twin weights was 52 ± 42%. The twin weight ratio was more than 70% in 25% of cases. When this characteristic was present, the incidence of preterm delivery was 90%, polyhydramnios 40%, and congestive cardiac failure in the pump twin 30%, as compared with 75%, 30%, and 10%, respectively, when the ratio was less than 70% (Moore et al. 1990).

In the series by Healey (1994), of 5 cases at Monash Medical Centre and a review of 184 case reports in the literature from 1960 to 1991, the overall perinatal mortality for the pump fetus was 35% in twins and 45% in triplets. Factors associated with a significant increase in perinatal mortality for the pump fetus included delivery before 32 weeks of gestation, the acardius anceps form of acardia, and the presence of arms, ears, larynx, trachea, pancreas, kidney, or small intestine in the acardiac fetus.

Another study questioned the poor prognosis associated with pregnancies complicated by TRAP and explored the role of expectant management (Sullivan 2003). Ten cases of antenatally diagnosed acardiac twins delivered between 1994 and 2001 in one community were evaluated. All cases were managed expectantly. Nine women delivered a healthy pump twin. There was one neonatal death. The mean gestational age at delivery was 34.2 weeks and the mean weights of the pump and acardiac twins were 2279 g and 1372 g, respectively. The authors concluded that neonatal mortality of pump twins in antenatally diagnosed acardiac twin pregnancies may be considerably less than reported, and expectant management with close antepartum surveillance may be an option.

We have been guided by the acardius-to-pump twin ratio in deciding which pregnancies can be managed expectantly and which require fetal intervention. If the acardius-to-pump twin ratio remains < 0.7 and the echocardiogram remains normal, we manage TRAP sequence expectantly. In Dr. Crombleholme’s personal experience of 30 cases of TRAP sequence with acardius-to-pump twin ratio which remained < 0.7 managed expectantly, none required fetal intervention, and all survived and delivered at a mean gestational age of 38 weeks. In addition, several TRAP sequence cases with very small acardius-to-pump twin ratios (< 0.4) were observed to spontaneously clot off the umbilical artery to the acardius due to the low flow state.

The goal of antepartum management of a pregnancy complicated by the TRAP sequence is to maximize outcomes for the structurally normal pump twin. Management of acardiac twin gestations is controversial. When the diagnosis is made, the gestational age should be documented by maternal history and standard biometric measurements of the pump fetus. The high and low-risk factors for perinatal mortality in the pump fetus must be evaluated through sonographic examination. In the absence of poor prognostic features (twin weight ratio > 0.70, elevated CVO, increased C:T ratio, congestive cardiac failure, polyhydramnios), expectant management with serial sonographic evaluation is reasonable (Malone and D’Alton 2000). Additional factors that place the pregnancy at high risk for perinatal mortality include features of acardius anceps demonstrating the presence of arms, ears, larynx, trachea, pancreas, renal tissue, and small intestine. Rapid growth of the acardiac twin may also be a sign of poor outcome (Brassard et al. 1999).

Features that indicate a lower risk include features of acardius amorphous with the absence of arms, legs, brain, esophagus, trachea, and omphalocele (Healey 1994). Karyotyping of the pump twin should be offered because as many as 9% of pump twins have an abnormal karyotype (Healey 1994). Steroids should be given if delivery is expected between 24 and 34 weeks of gestation (NIH Consensus Development Panel 1995). Preterm labor should be suppressed with tocolytic agents.

Delivery at a tertiary-care hospital is recommended because of the risk of preterm delivery and congestive cardiac failure in the pump twin. The vaginal route is the preferred mode of delivery. The indications for cesarean include the standard obstetric reasons. In Moore et al.’s series, abnormal presentation and fetal distress necessitated cesarean delivery in more than half of the potentially viable pregnancies.

Medical management with maternal administration of digoxin or indomethacin has been anecdotally reported, but there are no case series using these management strategies. The use of maternal digitalization to treat cardiac failure in the pump twin was reported by Simpson et al. in 1983. Marked edema of the trunk in the normal twin was present. Fetal ascites, pleural effusion, or cardiomegaly was not demonstrated. Serial ultrasound examinations demonstrated the resolution of the edema and continued normal growth of the viable fetus. Delivered at 34 weeks, the normal twin weighed 1860 g. The acardiac twin weighed 1810 g. No subsequent reports of digoxin therapy for acardia have been reported.

Ash et al. (1990) reported the use of indomethacin in an acardiac pregnancy complicated by polyhydramnios at 21 weeks as a means of reducing renal perfusion and amniotic fluid production. No evidence of cardiac failure was visualized in the pump twin. Indomethacin, 50 mg daily, was given to treat the symptomatic polyhydramnios because of the high risk of premature labor. The indomethacin was continued for 8.5 weeks. Oligohydramnios at 34 weeks prompted induction of labor, and spontaneous vaginal delivery occurred. The normal twin weighed 1865 g at birth, and the acardiac twin weighed 785 g (Ash et al. 1990). Many invasive procedures have been described with the goal of interrupting the umbilical circulation of the acardiac twin. There has been controversy in the literature concerning which cases are candidates for such procedures. Previously, it was recommended that invasive procedures be performed only after heart failure has developed (Platt et al. 1983). Some have recommended surgical intervention only after medical therapy has failed (Ash et al. 1990). Others consider the diagnosis of TRAP an indication for fetal intervention (Tsao et al. 2002).

Various percutaneous procedures have been described to interrupt the umbilical circulation in acardiac twins, including (1) insertion of a thrombogenic coil into the recipient twin’s umbilical cord; (2) injection of silk soaked in alcohol into the cord; (3) injection of absolute alcohol into the cord; (4) fetoscopic ligation of the acardiac fetus’s cord; (6) bipolar forceps cautery of the acardiac fetus’s cord; (7) thermocoagulation of the aorta of the acradiac fetus; (8) intrafetal radiofrequency thermablation; (9) intrafetal microwave ablation; and (10) high intensity focused ultrasound ablation (Porreco et al. 1991; Holzgreve et al. 1994; Quintero et al. 1994; Sepulveda et al. 1995; Arias et al. 1998; Rodeck et al. 1998; Challis et al. 1999; Tsao et al. 2002; Livingston et al. 2007, Stephenson et al 2015, Omemura 2013).

Injection of coils, or slerosants, is generally no longer performed because of the unreliability in achieving complete occlusion. Fetoscopic cord ligation may be associated with a failure rate of 10% together with a 30% risk of preterm rupture of membranes (Challis et al. 1999). Laser and cautery options have the advantage of generally requiring one access port in the uterus and therefore may be associated with less morbidity.

Robie et al. (1989) reported a case of selective delivery by hysterotomy of an acardiac acephalic twin fetus at 22.5 weeks of gestation with the subsequent delivery of the normal twin at 33 weeks of gestation. Fries et al. (1992) subsequently reported 5 cases of selective delivery in 1992. In one case, placental abruption occurred shortly after the procedure, leading to fetal death. Two cases were delivered at 35 weeks of gestation, and the remaining 2 delivered at 27 and 28 weeks.

Porreco et al. (1991) described the insertion of a helical metal coil under sonographic guidance to induce thrombosis in the umbilical artery of the acardiac twin at 24 weeks. The co-twin delivered at 39 weeks and had a normal course.

Quintero et al. (1994) described a percutaneous fetoscopic procedure that treated this condition at 19 weeks of gestation and was followed by the birth of a normal twin at 36 weeks of gestation. A further case was reported by McCurdy et al. (1993). A trial of maternal digoxin administration failed and was followed by a fetoscopic ligation of the acardiac twin’s cord at 19 weeks. Ultrasound examination on the first postoperative day indicated the death of the pump twin.

Holzgreve et al. (1994) injected multiple pieces of silk suture soaked in 96% alcohol into the umbilical cord of an acardiac twin at 21 weeks of gestation. This resulted in immediate interruption of flow in the cord and the ultimate delivery at term of a 2780-g healthy newborn. The advantage of this approach in comparison to umbilical cord ligation is the use of a much thinner needle. Less operative time is required, and there is no need for general anesthesia (Holzgreve et al. 1994).

Other methods of interrupting the circulation in the acardiac twin involve direct coagulation of the umbilical vessels or the aorta, using either laser photocoagulation or diathermy themocoaglulation. Laser photocoagulation of umbilical vessels using a neodymium yttrium aluminum garnet laser has been successfully reported, although this approach appears less likely to be successful when performed after 24 weeks gestation (Arias et al. 1998). This may be because umbilical vessels are too large to adequately photocoagulate when the gestational age is greater than 24 weeks. Thermocoagulation of the aorta of the acardiac fetus using diathermy via a wire passed through an 18-gauge needle has been successfully reported in four cases at 24 weeks gestation or less (Rodeck et al. 1998). The advantages of this latter approach include avoiding the need for micro-endoscopic instruments or skills, and avoiding the difficulties in identifying the target umbilical cord.

Intrafetal radiofrequency ablation (RFA) has also been utilized in cases of TRAP. RFA causes thermal injury with high-frequency radiowaves that denature proteins and initiate cell death through coagulative necrosis. In a series of 23 pregnancies complicated by TRAP and managed with RFA, there was a 91% survival rate with a mean gestational age of 35 weeks at delivery (Lee et al 2004). Livingston and Crombleholme et al. reported a 95% survival rate with an ultrasound-guided technique using a 17-gauge radiofrequency LeVeen needle (Boston Scientific) with a mean gestational age at delivery of 36 weeks in a series of 26 patients. Survival rates of 85% have been reported with fetoscopic cord coagulation likely as a consequence of two ports being required (Lewi et al. 2003). More recently, Dr. Crombleholme’s experience with radiofrequency ablation for TRAP sequence in 54 fetuses—in which the acardius-to-pump twin ratio exceeded 0.7, there was evidence of increased combined ventricular output, or polyhydramnios in the pump twin—was successful with 97.3% pump twin survival and delivery at a mean gestational age of 36.5 weeks. It is important to note that the success rate for radiofrequency ablation for other indications such as anomalous co-twin is significantly lower at 85%. We suspect that this is due to the high rate of blood flow in the vessels of anomalous co-twins compared to an acardius. The higher blood flow dissipates the heat generated by the RFA needs and makes it harder to coagulate. Conversely, in TRAP sequence, the reversed blood flow in the typically single umbilical artery cord of the acardius is usually sluggish and less able to dissipate the heat generated by the RFA needles and coagulation is more efficient.

The timing of fetal intervention in TRAP sequence remains controversial. Older data suggests that the acardius-to-pump twin ratio greater than 70% predicts a 90% pregnancy complication rate combining fetal loss and/or severe premature delivery. Lewi et al, have reported that there is a 33% rate of fetal demise of the pump twin by 18 weeks in TRAP managed expectantly (Lewi et al 2010). This raised the question of treating TRAP sequence prophylactically without waiting for the pregnancy to meet the criteria for intervention. Our own experience with TRAP has shown that with appropriate surveillance in the absence of an acardius-to-pump twin ratio > 70%, that fetal loss, polyhydramnios, and preterm delivery are rare events. Aiken et al, however, have proposed prophylactic treatment of acardius at the time of diagnosis even when less than 16 weeks’ gestation. There have been now at least 3 case reports of TRAP sequence being treated prophylactically at < 16 weeks (Aikin et al 2014, Cabassa et al 2014, Paramasivam et al 2010). Usually, this has been by intravital radiofrequency ablation, but interstitial laser photocoagulation has also been reported in among these early gestation cases (Pagani et al 2014). In a series of 17 cases treated between 15 and 18 weeks gestation with interstitial laser, Pagani reported an 82% pump twin survival with two cases having to be re-treated due to recanalization and persistent blood flow to the acardius. In this series, cases managed expectantly had an intrauterine fetal demise before 15 weeks’ gestation suggesting the early diagnosis of TRAP may identify a subset of TRAP pregnancies at greater risk (Pagani et al 2014) .

The approach at Connecticut Children’s Fetal Care Center is to offer radiofrequency ablation in cases of TRAP sequence in which the acardius-to-pump twin ratio > 70%, there is polyhydramnios, or elevated combined ventricular output (>600 ml/kg/min). Using these criteria, we have had a 98% pump twin survival rate with radiofrequency ablation. In addition, using these criteria, we have not had a pump twin loss during expectant management who did not meet these criteria. It is important to note that the majority of our patients present at 18 weeks or later, and it may well be that the natural history of TRAP presenting < 18 weeks could be different. There is insufficient data to make recommendations about fetal intervention this early without meeting criteria, but prophylactic intervention may be justified in cases of early gestation presentation.

It has been suggested that successful interruption of the acardiac circulation after 24 weeks gestation may require a more invasive approach, such as fetoscopic ligation of the umbilical cord (Arias et al. 1998; NcCurdy et al. 1993; Quintero et al. 1994). However, the series reported both by Tsao and Livingston included patients successfully treated after 24 weeks’ gestation.

A neonatologist should attend the delivery, especially in cases of untreated TRAP sequence. In Moore et al.’s (1990) series, admission to a newborn intensive care unit was required in 41% of the pregnancies and 59% of those reaching viability. Five of 29 live-born pump twins died during the newborn period. There is little information in the literature on the neonatal course of the pump twin. The main problems for the pump twin include complications of prematurity and congestive heart failure (Moore et al. 1990; Van Allen et al. 1983). Other frequent neonatal findings include massive hepatosplenomegaly, ascites with hypoplasia of abdominal musculature, edema, and hypoalbuminemia due to inadequate liver synthesis of albumin (Van Allen et al. 1983). Respiratory assistance as well as support of myocardial function with inotropic medication may be required. Early administration of surfactant therapy is indicated when premature delivery at less than 30 weeks of gestation is anticipated. Postnatal consultation with a pediatric cardiologist and echocardiography are recommended.

There is little information in the literature concerning long-term outcomes for the pump twin. In the series by Ozowa et al., long-term neurodevelopmental outcomes were studied in 38 patients who underwent RFA for TRAP sequence with a median age at assessment of two years five months. Children who survived TRAP sequence had uniformly favorable neurodevelopmental outcomes (Ozowa 2021). Considerations for the long-term prognosis must include the degree of prematurity, the severity of the neonatal course, and the degree of congestive heart failure.

Estimates of the recurrence risk of acardiac twin pregnancy are on the order of 1 in 10,000 (Van Allen et al. 1983). This recurrence risk is calculated from the recurrence risk for monoamniotic twinning, which is 1% (Myrianthopoulos 1970), multiplied by the frequency of the occurrence of the TRAP sequence, which is approximately 1% of all monozygous twins (Gillim and Hendricks 1953; Napolitani and Schreiber 1960).

Ahlfeld F. Die Entstehung der Acardiaci. Arch Gynakol 1879;14:321.

Aitken, K, Andrews J, VanMieghem T, Windrim R, Kachura J, Ryan G: Early radiofrequency ablation for twin reversed arterial perfusion sequence: Case report and literature review. Gynecol Obstet 2014, 4:9-1-5

Allan LD. Manual of fetal echocardiography. Boston: MTP 1986.

Arias F, Sunderji S, Gimpleson R, et al. Treatment of acardiac twinning. Obstet Gynecol 1998;91:818–821.

Ash K, Harman CR, Gritter H. TRAP sequence—successful outcome with indomethacin treatment. Obstet Gynecol 1990;76:960.

Bieber FP, Nance WE, Morton CC, et al. Genetic studies of an acardiac monster: evidence of polar body twinning in man. Science 1981;213:755.

Benirschke K, Kim CK. Multiple pregnancy. Part I. N Engl J Med 1973;288:1276–1284.

Benson CB, Bieber FR, Genest DR, et al. Doppler demonstration of reversed umbilical blood flow in an acardiac twin. J Clin Ultrasound 1989;17:291–295.

Berg C, Hoist D, Mallmann MR, Gottschalk I, Gembruch U et al: Early vs late intervention in win reversed arterial perfusion sequence. Ultrasound Obstet Gynecol 2014, 43 :60-64

Billah KL, Shah K, Odwin C. Ultrasonic diagnosis and management of acardius acephalus twin pregnancy. Med Ultrasound 1984;8:108.

Brassard M, Fouron JC, Leduc l, et al. Prognostic markers in twin pregnancies with an acardiacfetus. Obstet Gynecol 1999; 94: 409-414.

Cardwell MS. The acardiac twin, a case report. J Reprod Med 1988;33:320–322.

Cabassa P, Fishera VA, Prefumo F, Taddei F, Gandolfi S et al: The use of radiofrequency ablation for twin-reversed arterial perfusion sequence: A case series and review of the literature. Eur J Obstet Gynecol Reprod Biol 2013, 166: 127-132

Challis D, Gratacos E, Deprest JA: Cord occlusion techniques for selective termination in monochorionic twins. J Perinat Med 27(5):327-338, 1999.

D’Alton ME, Simpson LL. Syndromes in twins. In: Chervenak FA, D’Alton ME, eds. Multiple gestation. Semin Perinatol 1995;19:375–386.

Das K. Acardius anceps. Br J Obstet Gynaecol 1902;2:341.

Dashe JS, Fernandez CO, Twickler DM. Utility of Doppler velocimetry in predicting outcome in twin reversed-arterial perfusion sequence. Am J Obstet Gynecol 2001; 185: 135-139.

DeVore GR. Cardiac imaging. In: Sabbagha RE, ed. Diagnostic ultrasound. 2nd ed. Philadelphia: Lippincott 1987: 324–362.

Donnenfeld AE, van de Woestijne J, Craparo F, et al. The normal fetus of an acardiac twin pregnancy: perinatal management based on echocardiographic and sonographic evaluation. Prenat Diagn 1991;11: 235–244.

Fisk NM, Ware M, Stanier P, et al. Molecular genetic etiology of twin reversed arterial perfusion sequence. Am J Obstet Gynecol 1996;174:891–894.

Fries MH, Goldberg JD, Golbus MS. Treatment of acardiac-acephalus twin gestations by hysterotomy and selective delivery. Obstet Gynecol 1992;79:601–604.

Gillim DL, Hendricks CH. Holocardius: review of the literature and case report. Obstet Gynecol 1953;2:647.

Guimaraes CV, Kline-Fath BM, Linam LE, Calvo Garcia MA, Rubio EI, Lim FY: MRI findings in multifetal pregnancies complicated by twin reversed arterial perfusion sequence (TRAP). Pediatr Radiol 2011; 41: 694-701

Haring DAJP, Cornel MC, Van der Linden JC, et al. Acardius acephalus after induced ovulation: a case report. Teratology 1993;47:257–262.

Hata N, Wada T, Kashima K, Okada Y, Unno N, Kitagawa M, Chiba T: Non-gated fetal MRI of umbilical blood flow in an acardiac twin. Pediatr Radiol 2005; 35:826-829

Healey MG. Acardia: predictive risk factors for the co-twin’s survival. Teratology 1994;50:205–213.

Holzgreve W, Tercanli S, Krings W, et al. A simpler technique for umbilical-cord blockade of an acardiac twin. N Engl J Med 1994;331:56–57.

James WH. A note on the epidemiology of acardiac monsters. Teratology 1977;16:211–216.

Kappelman MD. Acardius amorphus. Am J Obstet Gynecol 1944;47:412.

Kinsel-Ziter ML, Cnota JF, Crombleholme TM, et al. Twin reversed arterial perfusion sequence: pre- and post-operative cardiovascular findings in the pump twin. Ultrasound in Obstet Gynecol 2009, in press.

Lachman R, McNabb M, Furmanski M, et al. The acardiac monster. Eur J Pediatr 1980;134:195.

Langlotz H, Sauerbrei E, Murray S. Transvaginal Doppler sonographic diagnosis of an acardiac twin at 12 weeks gestation. J Ultrasound Med 1991;10:175–179.

Lewi L, Valencia C, Gonzalez E, Deprest J, Nicolaides KH: the outcome of twin reversed arterial perfusion sequence diagnosed in the first trimester. Am J Obstet Gynecol 2010: 203: 213-218

Lee H, Wagner A, Ball R, et al. Radiofrequency ablation for TRAP sequence. Am J Obstet Gynecol2004; 191: S18.

Lee H, Bebbington M, Crombleholme TM: The North American Fetal Therapy Network Registry Data on outcomes of radiofrequency ablation for twin-reversed arterial perfusion sequence. Fetal Diagn Ther 2013, 33: 224-229

Lewi L, Gratacos E, Ortibus E, et al. Pregnancy and infant outcome of 80 consecutive cord coagulation in complicated monochorionic multiple pregnancies. Am J Obstet Gynecol 2006;194:782-789.

Livingston JC, Lim FY, Polzin W, et al. Intrafetal radiofrequency ablation for twin reversed arterial perfusion (TRAP): a single-center experience. Am J Obstet Gynecol 2007;197:399.e1-3.

Mack LA, Gravett MG, Rumack CM, et al. Antenatal ultrasonic evaluation of acardiac monsters. J Ultrasound Med 1982;1:13–18.

Malone FD, D’Alton ME. Anomalies peculiar to multiple gestations. Clin Perinatol 2000; 27: 1033-1046.

McCurdy CM Jr, Childers JM, Seeds JW. Ligation of the umbilical cord of an acardiac-acephalus twin with an endoscopic intrauterine technique. Obstet Gynecol 1993; 82:708–711.

Moore TR, Gale S, Benirschke K. Perinatal outcome of forty-nine pregnancies complicated by acardiac twinning. Am J Obstet Gynecol 1990;163:907–912.

Myrianthopoulos NC. An epidemiologic survey of twins in a large, prospectively studied population. Am J Hum Genet 1970;22:662.

Napolitani FD, Schreiber I. The acardiac monster: a review of the world literature and presentation of 2 cases. Am J Obstet Gynecol 1960;80:582.

Ozawa K, Wada S, Muromoto J, Sugibayashi R, Wada YS, Ito Y, Sago H: Longterm neurodevelopmental outcomes of pump twin in twin reversed arterial perfusion sequence treated by radiofrequency ablation. Prenatal Diagnosis 2021; 41: 1575-1581

Pagani G, D’Antonio F, Khalil A, Papegeorghiou A, Bhide A et al: Intrafetal laser for twin reversed arterial perfusion sequence: cohort study and meta-analysis. Ultrasound Obstet Gynecol 2013, 42: 6-14

Platt LD, DeVore GR, Bieniarz A, et al. Antenatal diagnosis of acephalus acardia: a proposed management scheme. Am J Obstet Gynecol 1983;146:857.

Porreco RP, Barton SM, Haverkamp AD. Occlusion of umbilical artery in acardiac, acephalic twin. Lancet 1991; 337:326–328.

Pretorius DH, Leopold GR, Moore TR, et al. Acardiac twin: report of Doppler sonography. J Ultrasound Med 1988; 7:413.

Quintero RA, Rich H, Puder K, et al. Umbilical-cord ligation of an acardiac twin by fetoscopy at 19 weeks gestation. N Engl J Med 1994;330:469–471.

Robie GJ, Payne GG, Morgan MA. Selective delivery of an acardiac, acephalic twin. N Engl J Med 1989;320:512.

Rodeck C, Deans A, Jauniaux E. Thermocoagulation for the early treatment of pregnancy with an acardiac twin. N Engl J Med 1998;339:1293–1295.

Schatz CF. Die Acardii und ihre Verwandten. Berlin: A. Hirschwald, 1898.

Shenker L, Reed KL, Marx GR, et al. Fetal cardiac Doppler flow studies in prenatal diagnosis of heart disease. Am J Obstet Gynecol 1988;158:1267–1273.

Sherer DM, Armstrong B, Shah YG, et al. Prenatal sonographic diagnosis, Doppler velocimetric umbilical cord study, and subsequent management of an acardiac twin pregnancy. Obstet Gynecol 1989;74:472–475.

Silverman NH, Kleinman CS, Rudolph AM, et al. Fetal atrioventricular valve insufficiency associated with nonimmune hydrops: a two dimensional echocardiographic and pulse Doppler ultrasound study. Circulation 1985;72:825–831.

Simonds JP, Gowen GA. Fetus amorphus: report of a case. Surg Gynecol Obstet 1925;41:171.

Simpson PC, Trudinger BJ, Walker A, et al. The intrauterine treatment of fetal cardiac failure in a twin pregnancy with an acardiac, acephalic monster. Am J Obstet Gynecol 1983;147:842–844.

Sogaard K, Skibsted L, Brocks V. Acardiac Twins: pathophysiology, diagnosis, outcome, and treatment. Fetal Diagn Ther 1999; 14: 53-59.

Steffenson TS, Gilbert-Barness E, Spellacy W et al: Placental pathology in TRAP sequence: clinical and pathogenetic implications. Fetal and Pediatr Path 2008; 27:13-29

Sullivan AE, Varner MW, Ball RH, Jackson M, Silver RM. The management of acardiac twins: a conservative approach. Am J Obstet Gynecol 2003; 189: 1310-1313.

Tsao KJ, Feldstein V, Albanese CT et al: Selective reduction of acardiac twin by radiofrequency ablation. Am J Obstet Gynecol 187(3):635-640, 2002

Van Allen MI, Smith SW, Shepard TH. Twin reversed arterial perfusion (TRAP) sequence: a study of 14 twin pregnancies with acardius. Semin Perinatol 1983;7:285.