The Detection, Outcome, and Presentation of Twin-Twin Transfusion Syndrome in Monochorionic Diamniotic Twin Pregnancies Followed with a Protocol of Fortnightly Ultrasound Examination

• What is already known about the topic?

• If detected in time, fetoscopic laser surgery of the vascular anastomoses improves perinatal outcomes.

• Survival rates after fetoscopic laser surgery are better in stage I–II twin-twin transfusion syndrome than in stage III–IV.

• In 90% of cases, a protocol of fortnightly ultrasound scan detects twin-twin transfusion in time. In 10%, twin-twin transfusion syndrome (TTTS) presents with fetal demise, especially before 16 weeks or after 26 weeks.

• The survival of TTTS in an unselected cohort is lower than that after fetoscopic laser surgery.

• In stage III–IV, abnormal Doppler measurements precede the diagnosis of TTTSs, suggesting that the staging system does not reflect a progressive worsening of the disease, and more frequent follow-up may not result in an earlier stage at diagnosis.

Twin-twin transfusion syndrome (TTTS) is the most severe complication in monochorionic diamniotic (MCDA) twin pregnancies. TTTS occurs in about one in ten pregnancies, usually between 16 and 26 weeks [1]. Untreated, it is associated with high mortality and morbidity rates [2]. Fetoscopic laser photocoagulation (FLP) of the placental anastomoses is currently the best treatment. Compared to amnioreduction, FLP is associated with higher perinatal survival rates, birth at a more advanced gestational age, and better neurodevelopmental outcomes [3]. In selected cases, such as a major anomaly in one twin, an inaccessible vascular equator, or patient preference, selective reduction may be an alternative to FLP to save one twin [4]. Invasive fetal therapy is only useful in the absence of intrauterine fetal demise (IUFD), with intact membranes and before cervical dilation.

Increased surveillance may help diagnose TTTS in time. Therefore, international guidelines recommend fortnightly sonographic evaluations for all MCDA twin pregnancies, starting at 16 weeks and continuing until birth [5‒8]. However, there are no comparative studies such as fortnightly scheme’s efficacy, and this recommendation is supported by a few cohort studies only [9‒11]. Also, although the outcome after FLP is well documented with recent series reporting a 74% survival with at least one twin surviving in 88% [12], the outcome of TTTS in an unselected monochorionic twin population is not well known. Finally, common knowledge suggests that a more frequent follow-up may pick up the disease at an earlier stage, improving outcomes [13]. However, little is known on the ultrasound findings before the diagnosis.

The primary aim was to assess the proportion of TTTS pregnancies that is detected in time by a fortnightly ultrasound exam. We defined in time as before IUFD, ruptured membranes, and cervical dilation because we do not offer FLP under these circumstances. Second, we document the survival and neonatal outcome of TTTS in an unselected cohort of MCDA twin pregnancies followed from the first trimester and the ultrasound findings within 14 days of the diagnosis of TTTS.

We performed a retrospective study of ongoing MCDA twin pregnancies diagnosed in the first trimester (11 + 0–14 + 0 weeks) between January 2002 and September 2018. MCDA twin pregnancies are routinely referred to the twin clinic at the University Hospitals of Leuven (study center) for a detailed ultrasound scan at 11–14, 16, 20, and 26–30 weeks. In addition to these 4 examinations, patients are instructed to have a sonographic assessment at least every 2 weeks, either at the study center or the referring institution, to detect TTTS in a timely fashion as per recommendation [5‒8]. Patients explicitly referred in the first trimester for invasive testing or because of an anomaly were not included. The same cohort was included in an earlier study on selective growth restriction and the impact of in vitro fertilization [14, 15].

The ultrasound scans were performed according to a well-defined protocol [1, 14] by experienced sonographers on Voluson E10/E8/E6/730 (GE Healthcare, Chicago, IL, USA). Ultrasound reports were made using the Astraia software (Astraia software gmbh, Munich, Germany). All patients were informed about TTTS and its warning signals at the first visit. When we detected a discordance in growth, amniotic fluid, or Doppler findings, we increased the frequency of ultrasound evaluations to weekly or twice-weekly exams depending on the severity and the gestational age.

Patients with TTTS were offered FLP as first-line treatment. Selective reduction by umbilical cord coagulation or intra-fetal radiofrequency ablation was offered as an alternative, especially if one twin had an anomaly, was deemed to have a poor prognosis, or if FLP was technically not feasible. Patients could also opt for termination of pregnancy if TTTS presented before viability. After 28 weeks, TTTS was mostly managed either by an elective birth or by amnioreduction. The Ethics Committee of the University Hospital Leuven approved the study (S63561).

TTTS was diagnosed by oligohydramnios and a small or empty bladder in the donor and polyhydramnios with a distended bladder in the recipient. Oligohydramnios was defined as a deepest vertical pocket (DVP) less than 2 cm [3]. Polyhydramnios was defined as a DVP more than 6 cm before 16 weeks [16], more than 8 cm before 20 weeks, and more than 10 cm after 20 weeks [3]. TTTS severity was assessed based on the Quintero staging system [17]. Timely detection of TTTS was defined as a diagnosis before IUFD, ruptured membranes, or dilated cervix.

Survival refers to fetal and neonatal survival up to 28 days of life per total number of included twin fetuses in the first trimester. Fetal loss included IUFD, miscarriage, and termination of pregnancy. Neonatal demise refers to the number of deaths within the first 28 days per total number of live-born babies. Severe neonatal morbidity was defined as either the occurrence of chronic lung disease (defined as oxygen dependency at 36 weeks), patent ductus arteriosus needing surgical treatment, necrotizing enterocolitis grade II or higher, retinopathy of prematurity stage III or higher, ischemic limb injury and amniotic band syndrome, and severe cerebral injury. Severe cerebral injury was defined as intraventricular hemorrhage grade III or higher, cystic periventricular leukomalacia grade II or higher, ventricular dilatation greater than the 97th percentile, porencephalic or parenchymal cysts, or other severe cerebral lesions associated with adverse neurological outcome. Neonatal brain scans were not always performed in infants after 32 weeks. In these cases, we collected the results from the fetal brain magnetic resonance imaging (MRI), which we have been offering to all TTTS pregnancies at 28–32 weeks from 2010 onward.

Ultrasound data were retrieved from the Astraia database. Pregnancy and neonatal outcome data were collected after birth from the electronic clinical records. If not present, they were requested from the referring physicians.

The primary outcome is the proportion of TTTS pregnancies detected in a timely fashion. The secondary outcomes are the fetal-neonatal survival and neonatal morbidity according to the management strategy, and the findings on the last scan prior to the diagnosis of TTTS in the subpopulation of patients that received this scan in our center.

During the study period, 678 MCDA pregnancies were followed from the first trimester onward. Three patients were lost to follow-up, which left 675 in the analysis. The average maternal age was 30 years, and 316/675 (47%) were nulliparous. In 573/675 (85%), the pregnancy was conceived spontaneously. The compliance with the 16, 20, and 30 week’s exam at the study center was 613/637 (96%), 564/583 (97%), and 463/544 (85%), respectively [14]. Of the 154 (23%) patients, who initially booked for antenatal care in our center, the median scan interval was 14 days (range 1–29 days). Excluding the 3 patients that already developed TTTS by 16 weeks, the scan interval never exceeded 21 days in 136 of 151 (90%) patients (Fig. 1).

TTTS occurred in 82 of 675 (12%) pregnancies. In 11 of 82 (13%), TTTS was diagnosed before 16 weeks and 12 of 82 (15%) after 26 weeks. The survival rate was 105/164 (64%), with the survival of at least one twin in 65/82 (79%). Severe neonatal morbidity occurred in 10/105 (10%), and severe cerebral injury was diagnosed in 6/105 (6%). In 86/105 (82%) infants, brain scans were performed in the neonatal period. From the remaining 19 infants, all born between 33 and 39 weeks, information on cerebral injury was obtained from the third-trimester MRI scan.

In 8/82 (10%) cases, the diagnosis of TTTS was made after IUFD of one or both twins (stage V). None presented with ruptured membranes or a dilated cervix at diagnosis. The clinical characteristics of stage V diagnoses are summarized in Table 1 (case 1–5) and Table 2 (case 6–8).

In 5/8 stage V patients, TTTS was diagnosed before 16 weeks (case 1–5). At the first-trimester scan, 4 had discordant amniotic fluid (case 1–4), 3 had a crown-rump length (CRL) discordance of more than 20% (case 1–3), and in another case, acrania was diagnosed in the larger twin (case 4). Therefore, these 4 pregnancies (case 1–4) had additional ultrasound scans 3 to 10 days before the demise.

In 3/8 stage V patients, TTTS was diagnosed after viability (case 6–8). One double IUFD occurred at 24 weeks in a patient who was followed alternating with the referring physician and did not attend her ultrasound appointments between 16 and 24 weeks (case 6). This same patient also had a stage V TTTS at 15 weeks in a subsequent pregnancy (case 5). One single IUFD (case 7) occurred at 26 weeks, 2 weeks after a normal exam, and a double IUFD (case 8) occurred at 27 weeks, 8 days after the ultrasound scan showed mild discordant fluid.

The survival in the 8 stage V cases was 2/16 (12.5%) with the survival of at least one twin in 2/8 (25%). One surviving twin underwent an intrauterine transfusion at 16 weeks. Both surviving twins had a normal fetal brain MRI at 30–32 weeks and were born at term.

In 74/82 (90%), TTTS was diagnosed before IUFD of one or both twins. The clinical details, management, and outcome of TTTS pregnancies that were detected in a timely fashion are summarized in Table 3.

In 48/74 (65%) patients, FLP was performed as a first-line intervention. The survival rate was 77/96 (80%), with the survival of at least one twin in 45/48 (94%). Four patients required an additional intervention: 3 selective reductions (2 recurrent TTTS and one severe cerebral injury) and one intrauterine transfusion/exchange transfusion for twin anemia polycythemia syndrome.

In 15/74 (20%) patients, a primary selective reduction was performed for a discordant anomaly (n = 6), severe growth restriction (n = 4), or technical limitations to perform an FLP procedure (n = 5). In 9 of 15 pregnancies (60%), selective reduction resulted in a surviving infant.

In 9/74 (12%) patients, no surgical intervention was performed, resulting in a survival rate of 17/18 (94%) and the survival of at least one twin in 9/9 (100%). Four patients were electively delivered between 30 and 33 weeks. Two other patients underwent an amnioreduction at 28 and 30 weeks. Another 3 patients were managed expectantly. One case was diagnosed at 15 weeks (stage III), the second at 16 weeks (stage II), and the third at 20 weeks in stage I. Because of technical limitations, we chose initially for conservative management in these 3 patients, which resulted in a spontaneous resolution, albeit with the unexpected IUFD of a former recipient at 26 weeks in the third case.

Finally, 2/74 (3%) patients with stage II and III disease opted for termination as primary management. In one pregnancy, both twins had a congenital heart defect.

Two-thirds of TTTS pregnancies (55/82) had an ultrasound scan at the study center within 14 days of diagnosis with a median interval of 4 days. The ultrasound findings on the pre-TTTS scan of pregnancies presenting as mild (stage I and II) and severe (stage III and IV) TTTS are summarized in Table 4. Of the patients who presented as severe TTTS, the majority already had abnormal Doppler measurements before TTTS was diagnosed. There was no difference in staging and outcome between pregnancies that did or did not have their previous scan at the study center within 14 days of diagnosis (Table A1).

This study demonstrates that a protocol of fortnightly sonographic follow-up detects 9 out of ten TTTS cases in a timely fashion. As such, TTTS was diagnosed in stage V in 8 of 675 (1%) MCDA twins. This high rate of timely detection is in agreement with the results of Sueters et al. [9], who reported the timely detection of 4 TTTS cases in 23 patients using a similar schedule.

In our series, 5 of 8 stage V pregnancies presented before 16 weeks, mostly within 10 days of a previous exam. At the first-trimester scan, 4 had discordant amniotic fluid, 3 had a CRL discordance of more than 20%, and one had a nuchal translucency discordance of 65%. CRL and nuchal translucency discordance increase the risk of TTTS [18]. In a recent series reporting on 1,274 MCDA pregnancies, a CRL discordance of more than 20% was present in 1% of MCDA twins, but 70% were complicated by IUFD or needed FPL before 20 weeks [19]. Since FLP can only be performed after amnion-chorion fusion, these early transfusion imbalances may result in IUFD before FLP can be offered. At present, the benefit of FLP performed before 16 weeks remains uncertain [20]. Therefore, the clinical relevance and impact of detecting these early TTTS cases may be limited.

Two previously uncomplicated pregnancies suddenly developed polyhydramnios after 26 weeks and were diagnosed in stage V. The exact mechanism of these acute, late presentations is unknown but may be related to fetal vessel thromboses, which is more common in monochorionic placentas [21]. Such transfusion imbalances due to the occlusion of an anastomosis will remain unpredictable.

In our series, 49% of TTTS patients took home both babies, which is lower than the 67% dual survival rate after FLP in this study, and the 62% dual survival reported in the recent literature [12]. In our series, only 48/82 (59%) of TTTS cases were treated with FPL, which explains why the actual mortality of TTTS is higher than what is reported after FPL.

Of the patients presenting as stage III–IV TTTS, most already had abnormal Doppler findings on their pre-TTTS scan. Therefore, these patients did not pass through stage I and II disease first. Severe TTTS does not appear to result from a delayed diagnosis but seems to represent a different presentation of the disease in which Doppler abnormalities precede a severe amniotic fluid discordance. Therefore, the Quintero’s stages of TTTS should not be interpreted as a consecutive worsening of the disease, but rather as distinct presentations with different underlying pathophysiology. Also, as Doppler abnormalities occur before a severe amniotic fluid discordance, a more frequent follow-up will not necessarily prevent a severe stage at diagnosis. Similarly, other studies did not find a relationship between the follow-up interval and staging at first diagnosis [10, 11].

Our study’s strength is that it reports on an unselected cohort of MCDA twin pregnancies followed from the first trimester with fortnightly follow-up. As a limitation, we do not have exact information on the fortnightly scheme’s compliance in patients who were followed up alternating with the referring physician. Indeed, one such patient with stage V TTTS at 24 weeks did not attend her appointments after 16 weeks. In 90% of patients who booked at our center, the scan interval never exceeded 21 days. Also, scan attendance at 16, 20, and 30 weeks was 96, 97, and 85%, respectively. In Belgium, ultrasound appointments are covered by national health insurance, and patients are keen to know how their twin pregnancy evolves. Therefore, we expect a similarly good adherence for patients who were followed alternating with their referring physician.

In conclusion, our study shows that a scheme of fortnightly ultrasound scans is effective in detecting TTTS in time. None presented with ruptured membranes or cervical dilation, and most stage V cases were diagnosed before 16 weeks or presented with acute TTTS after 26 weeks. Only about 60% of all TTTS cases were treated with FLP. Therefore, the mortality was higher than that reported after FLP. Finally, most patients had abnormal Doppler findings before the diagnosis of stage III–IV TTTS, suggesting that these pregnancies did not pass through stage I and III first.

We thank the physicians for referring their patients for alternating follow-up in our center.

The research was conducted ethically in accordance with the World Medical Association Declaration of Helsinki. The study was approved by the Medical Ethics Committee of the University Hospital Leuven, Leuven, Belgium (S62385). The local Ethics Committee of the University Hospital Leuven agreed on using retrospective data collected by the therapeutic team without written informed consent of included patients.

The authors have no conflicts of interest to declare. Liesbeth Lewi and Jan Deprest are both Editorial Board Members of Fetal Diagnosis and Therapy.

This study was funded by a grant from the Fund for Academic Research of the University Hospitals Leuven, Belgium; LL is the recipient of a grant of “Fonds voor Wetenschappelijk Onderzoek” (FWO grantnr: 1804718N).

Study conception and design: Isabel Couck, Sophie Ponnet, and Liesbeth Lewi; data acquisition: all authors; analysis and data interpretation: Isabel Couck, Sophie Ponnet, and Liesbeth Lewi; drafting of the manuscript: Isabel Couck, Sophie Ponnet, and Liesbeth Lewi; and critical revision: all authors. Isabel Couck and Sophie Ponnet contributed equally to this paper.