Case

A 29-year-old woman was referred to our center due to triplet pregnancy at 13 weeks of gestation. One fetus was diagnosed as an acardius, and blood flow was detected from the other two twins, the so-called twin reversed arterial perfusion (TRAP) sequence. Both normal twins showed a normal amniotic fluid pocket, and Doppler ultrasound did not show any abnormality. However, weekly examinations showed growth of the acardius, and the reversed arterial blood flow persisted. To prevent anemia and death of the pump twins subsequently caused by sudden blood flow arrest of the acardius [1], we scheduled blood flow coagulation of the acardius by radiofrequency ablation (RFA) [2]. Preoperatively, we performed a dual-gate Doppler examination with a sector C35 probe (2–8 MHz) connected to an ARIETTA 70 ultrasound machine (Hitachi-Aloka Medical, Ltd., Tokyo, Japan); this could identify which fetus was more responsible for the acardius circulation. First, we theorized that the proximal cord insertion might be the artery feeding the acardius, but there was no synchronicity of the arterial pulse rate (Fig. 1, upper row). Then, we measured the other distal site of the cord umbilical arterial pulse wave, and the heart rate completely synchronized (Fig. 1, lower row). After several examinations, we confirmed the initial findings. These findings suggested that the distal site circulation would be from the primary pump twin whose arteries connected with the acardius; this identification clarified the physiological information of the main communicating vessels for managing the monochorionic-triamniotic triplets (Fig. 2). At 16 weeks 3 days of gestation, we coagulated the acardius’ pelvic blood flow by ultrasound-guided RFA under local anesthesia. We succeeded in coagulation and closure of the abnormal communication to the acardius. Postoperatively, neither twin showed any signs of abnormal circulation such as anemia, and the patient resumed outpatient prenatal care. At 29 weeks 3 days, preterm rupture of the membranes followed by labor occurred; we performed cesarean section and delivered 1167/1237-g female infants and a macerated acardius. They were managed in the NICU, where they stabilized without any sequelae. The placenta was found to have two superficial anastomoses, thick artery–artery and vein–vein anastomoses communicating from the distal insertion site, and we determined that these substantial communications were responsible for the TRAP sequence (Fig. 3; Nos. 1–3). A tiny, single communicating vessel was also identified from No. 2 to No. 3. This vessel did not have any return vessels; therefore, it might not have been a feeding communication to the acardius.

Fig. 1
figure 1

Upper row a dual-gate pulsed Doppler image of cord insertion of the acardius (No. 3) and the proximal fetus (No. 2). There is no synchronicity of the arterial pulse; thus, we theorized that this No. 2 fetus had no significant circulatory connection to the acardius and was not the pump twin. Lower row a dual-gate pulsed Doppler image of the cord insertion of the acardius (No. 3) and distal fetus (No. 1). There was synchronicity of the arterial pulse; thus, we theorized that this No. 1 fetus had the primary circulatory connection to the acardius and was responsible for the development of the TRAP sequence

Fig. 2
figure 2

A schematic of our hypothesis after conventional and dual-gate Doppler examinations. We clarified the communication of TRAP sequence No. 1 (pump twin) to No. 3 (acardius) even though the insertion was further from the other twin (No. 2)

Fig. 3
figure 3

The placenta connected to two thick blood vessels that connected the pump twin (No. 1) to the acardius (No. 3): an artery–artery anastomosis and a vein–vein anastomosis. A tiny, single communicating vessel was also identified from No. 2 to No. 3; this vessel had no return vessels from the acardius to No. 2

Discussion

Dual-gate Doppler is useful for evaluating two pulse waves in real time with completely zero intra-observer error [3]. Since we used it for a TRAP sequence with two possible pump twins detected by conventional Doppler and B-mode examination, this synchronicity was very useful for identification of the pump twin. Some authors reported on the use of this procedure for two vessel measurements in one fetus.1 However, its usefulness for identification of two different fetal circulations has not been reported to date. Thus, this would be the first report of identification of the primary pump twin in a TRAP sequence of a MT triplet pregnancy using one of the newest ultrasound technologies. Fetal therapy for TRAP sequence has been proposed within the past decade1 because many unexpected fetal deaths occur at ≤16 weeks of gestation. Currently, we perform fetal therapy in the preventative stage rather than waiting for the deteriorated stage. This policy takes into account the fact that conventional chronic signs of fetal cardiac stress such as deterioration of polyhydramnios or fetal Doppler flow abnormalities would not be markers for identification of the pump twin. If we could detect these risks before deterioration occurred, this would also contribute to focusing postoperative management on the most-affected fetus. Since the incidence of triplet TRAP sequence is rare, this new technology definitely launches a new field of fetal science.