Aspects of twin-twin transfusion syndrome (TTTS) diagnosis, treatment alternatives, and research opportunities were considered during a consensus conference that was held by the North American Fetal Therapy Network in 2009. A 3-member scientific consensus panel gathered data from expert conference presentations, postconference communications, and comprehensive scientific literature database searches to develop recommendations for TTTS diagnosis, therapy, and research. The panel recommends retaining the Quintero staging system until a superior system has been validated appropriately. It concludes that there is normative equipoise to justify the performance of randomized clinical trials to identify the optimal treatment strategy for mild TTTS. Recommendations for the design and conduct of clinical trials and observational studies are also provided.
This review represents a synthesis of current knowledge, experience, and expert opinion that has been gathered on the topic of mild twin-twin transfusion syndrome (TTTS). Aspects of TTTS diagnosis, treatment alternatives, and research opportunities were considered during a consensus conference that was held by the North American Fetal Therapy Network (NAFTNet) in February 2009. NAFTNet is a voluntary association of 20 medical centers in Canada and the United States that perform advanced in utero fetal therapeutic procedures. The missions of NAFTNet are to form a cooperative clinical research network to advance fetal therapy and improve neonatal outcomes and to provide education and training in fetal therapy. The purpose of the NAFTNet consensus conference was to evaluate the available scientific information regarding the management of Quintero stage I TTTS to develop consensus on the known benefits and risks of available therapies and to provide guidance on the most pertinent research goals and strategies to advance medical science on the evaluation and treatment of mild TTTS.
See related editorial, page 1
Based on existing controversies that were identified by international centers involved with TTTS care and research, NAFTNet charged the conference participants and the scientific consensus panel to address 3 major objectives: (1) consider revision or replacement of the Quintero staging system of TTTS to include additional physiologic parameters; (2) review what is known about the natural history of mild TTTS, how various treatments of TTTS stage I affect the progression to higher stages of disease, and evaluate the efficacy of laser photocoagulation therapy for stage I TTTS; and (3) determine whether there is equipoise to justify a randomized controlled trial for stage I TTTS.
The overall goal therefore was to attempt to develop a consensus for the appropriate management of Quintero stage I TTTS, and if consensus could not be established, determine whether there is sufficient equipoise to justify a prospective randomized trial for the treatment of mild or stage I TTTS.
To facilitate reaching an objective consensus, NAFTNet used a scientific consensus panel (T.R.M., W.D.F., D.M.S.) comprised of members who were not actively involved in TTTS assessment, care, or research and who had no financial conflict of interest. Additionally, NAFTNet conference organizers invited recognized experts in TTTS from the international community to allow a complete presentation of current data, controversial issues, and opinion regarding the consensus questions. Experts included obstetricians, maternal-fetal medicine specialists, pediatric surgeons, pediatricians, neonatologists, pathologists, and ethicists.
The scientific consensus panel gathered data from the conference presentations, postconference communications, and comprehensive scientific literature database searches to develop this consensus statement that addresses these 3 conference objectives and outlines recommendations for TTTS therapy and research.
Objective 1: evaluate the value of revising or replacing the Quintero staging system
Background
Although an uncommon complication of monochorionic twinning (10%), TTTS accounts for significant morbidity and death. TTTS occurs in monochorionic twin gestations because of unbalanced blood flow across placental anastomoses that connect the 2 placental circulations. If the net flow of blood between twins is excessive, the donor twin experiences decreased body growth, volume depletion, oliguria and oligohydramnios, and the recipient experiences volume overload, polyuria, polyhydramnios, and cardiac hypertrophy. Ultimately, 1 or both twins may become hydropic and/or die.
When recognized clinically, TTTS conveys a significant risk for fetal morbidity and death, particularly if untreated and if diagnosed at <28 gestational weeks. TTTS-induced death of at least 1 fetus in a twin pair is as high as 80-100% without treatment. Furthermore, death of 1 twin can result in subsequent death (12%) or neurologic damage (18%) of the co-twin. Although a number of treatments have been advanced for this condition, significant risks of morbidity and death remain. For example, with laser photocoagulation of placental vessels, preterm premature rupture of membranes within 1 and 3 weeks of the procedure have been reported to be 7% and 17%, respectively. Other complications that are less frequent include amniotic fluid leakage into the maternal peritoneal cavity (7%), vaginal bleeding (4%), abruption (2%), and chorioamnionitis (2%). For this reason, a rational selection of cases that are appropriate for invasive treatment is essential, especially considering that up to two-thirds of cases will remain stable or regress without therapy.
Because staging systems are used extensively and successfully in cardiovascular and oncology medicine, development of a reliable TTTS staging system offers significant opportunity to improve care. Typical goals of a disease staging system include (1) provision of a sequential description of progressively worsening aspects of disease; (2) identification of risk factors for progression of disease; (3) provision of referring medical providers with criteria for escalation of care to a specialized center or specialist; and (4) provision of a definitional framework by which to evaluate therapeutic trials.
Evidence that supports the correlation of the Quintero staging scheme and outcome
Quintero et al in 1999 proposed a 5-stage scheme to categorize TTTS cases with a hope of defining the prognostic value of sonographic and clinical parameters that could be used to predict which TTTS cases would proceed to fetal hydrops or death. The Quintero staging scheme uses discrete/categoric criteria (eg, oligohydramnios [yes/no], fetal bladder visible [yes/no]) as opposed to cutoffs within continuous variables. The original Quintero staging system is summarized in Table 1 .
| Quintero stage | Stage I | Stage II | Stage III | Stage IV | Stage V |
|---|---|---|---|---|---|
| Ultrasound finding | Maximum vertical pocket of amniotic fluid: donor <2 cm; recipient >8 cm | Nonvisualization of donor bladder over 60 min | Critically abnormal Doppler findings a | Hydrops in either twin | Death of either twin |
a Absent or reversed umbilical artery velocity, reversed ductus venosus velocity, pulsatile umbilical vein velocity.
Several issues should be considered when evaluating the usefulness of the Quintero staging system. Abnormalities of amniotic fluid volume have been recognized as a keystone sign of TTTS since the earliest reports. The choice of an 8-cm maximal vertical pocket of amniotic fluid on ultrasound scan as the threshold for polyhydramnios and a 2-cm pocket for oligohydramnios corresponds with the 90th and 10th percentiles, respectively. However, the relevance of these benchmarks in the prediction of adverse perinatal outcome is controversial. Similarly, visualization or nonvisualization of the bladder in the donor twin, which is a criterion for Quintero stage II, is not consistently representative of fetal physiology. Yamamoto et al, who measured urinary output with 3-dimensional ultrasound imaging, documented oliguria or anuria in 48% of donors with “visible” bladders assigned to stage I.
Subsequently in 2003, Quintero et al applied the staging system to 78 serial amnioreduction patients and 95 selective laser photocoagulation cases. Successful pregnancy outcome (defined as at least 1 surviving infant) correlated with stage in the serial amniocentesis group, but not in the laser photocoagulation group ( Figure 1 ). The lack of correlation with outcome and stage in the laser group could be explained by the fact that the circulatory source of the pathologic condition was eliminated by the laser treatment; with serial amnioreduction, the circulatory imbalance remained. Similar findings regarding the lack of correlation of stage and outcome for laser-treated cases were reported by Rossi and D’Addario.
Evidence relating Quintero staging to TTTS severity and progression
Dickinson et al assessed the value of Quintero stage in predicting the progression of TTTS using a series of 71 cases, of which 73% of the patients were treated with amnioreduction. In 21% of cases, TTTS resolved completely; disease resolution was associated with pregnancy prolongation, greater gestational age at delivery (36 vs 28.4 weeks; P < .001), and increased perinatal survival (100% vs 42.6%; P < .001). Logistic regression analysis was used to estimate that the probability of both infants surviving was 80% if the pregnancy remained at stage I or II, compared with 50% in stage III or IV at 26 weeks’ gestation, and only 25% if the disease reached stage III or more by 16 weeks’ gestation. As shown in Figure 2 , >60% of the cases that initially were assigned to stage I or II resolved or remained in their original stage, whereas less than one-third of the cases in stage III and none of the cases in stage IV returned to stage I or II. Further, when the gestational age at delivery was analyzed according to stage at diagnosis, there was a significant association with higher initial Quintero stage and earlier delivery ( P < .03; Figure 3 ).
It is remarkable that the present Quintero staging system can provide even approximate information regarding the risk of perinatal death in TTTS, given the highly complex physiologic conditions that are involved. Designing a single staging system to prognosticate the outcomes of cohabiting twins with differing placental resources and cardiovascular systems is a truly ambitious project. Therefore, and despite the encouraging correlations between the Quintero stages in nonlaser-treated subjects, reports of novel attempts to provide additional prognostic accuracy with the use of more complex imaging techniques have appeared recently.
Candidate criteria to be added to the Quintero TTTS staging scheme
Interesting but not rigorously validated or consistently reliable parameters that have been considered for TTTS severity prediction or prognosis include increased nuchal translucency in the recipient during first trimester, folding of the intertwin membrane at 15-17 weeks’ gestation, and intertwin hemoglobin discordance as indicated by middle cerebral artery Doppler interrogation for peak systolic velocity.
Conversely, placental vascular anastomoses, amniotic fluid/renal function variables, and cardiovascular performance factors have demonstrated significant promise for the improvement of prognostication in TTTS.
Artery-artery anastomoses
The finding of intertwin artery-to-artery (AA) anastomoses has been linked to improved outcome within a given TTTS stage. It is theorized that, although arteriovenous anatomoses promote 1-way blood flow and “twin transfusion,” AA anatomoses buffer the volume changes by equalizing blood pressure and volume (the recipient’s blood volume is raised through the arteriovenous connection but is modulated by AA anastomoses). The presence of AA anastomoses in addition to arteriovenous anastomoses has been associated with a higher survival rate than those pregnancies without AA anastomoses, independent of Quintero stage. In 1 series of monochorionic twins, TTTS occurred predominantly in the cases with arteriovenous anastomoses that lacked compensating superficial AA anastomoses ( P = .005). Thus, antenatal detection of AA anastomoses might be useful in triage of stage II cases to observation rather than invasive therapy.
Unfortunately, sonographic identification of AA anastomoses can be difficult. Taylor et al compared color Doppler images with placental pathologic findings in 105 monochorionic twins. AA anastomoses were identified with ultrasound imaging in 56% of cases and on pathologic examination in 65% cases, which gave a sensitivity and specificity of 85% and 97.3%, respectively. However, the best sensitivity was skewed to later gestations in which the usefulness of this finding was diminished.
Amniotic fluid volume and donor bladder size in stage I disease
The predictive value of amniotic fluid volume and donor bladder size for TTTS progression was assessed by O’Donoghue et al in 46 cases of stage I TTTS. In most cases, disease remained stable (28%) or resolved (41%). Of cases that progressed, 79% did so within 2 weeks; 93% progressed to at least stage III. Amniotic fluid, which was assessed by maximal vertical pocket or amniotic fluid index, was not predictive of progression nor was a “small” fetal bladder. Interestingly, there was a trend toward more cases with absent AA anastomosis among those infants who progressed ( Table 2 ).
| Outcome | n | Median gestational age at diagnosis, wk | AFI, cm | Median recipient MVP, cm | Patients with AAA, % | Patients with small donor bladder, % | Weight discordance >25% , % |
|---|---|---|---|---|---|---|---|
| Progression | 14 | 19.7 | 36.9 | 10.9 | 14.3 | 57.1 | 28.6 |
| Regression/stable | 32 | 21.6 | 34.1 | 9.9 | 37.5 | 25.0 | 31.2 |
Sonographic indices of myocardial function
During the volume loading stress of the recipient twin in TTTS, systolic and diastolic dysfunction frequently coexist. Tei developed a combined measure of ventricular function (now called the Myocardial Performance Index [MPI]) using sonographic evaluation of the cardiac cycle: MPI = (ICT + IRT)/ET, where ICT is isovolumic contraction time, IRT is isovolumic relaxation time, and ET is ejection time ( Figure 4 ). This index is easily obtainable in adults, is reproducible, and has a narrow range in normal adults (mean ± standard deviation, 0.39 ± 0.05). The index is elevated in myopathic hearts, typically >0.7.
Because a worrisome fraction of stage I TTTS cases can transition rapidly to hydrops or fetal death, investigators have evaluated fetal cardiac indices in early stage TTTS. Michelfelder et al performed a retrospective study of 42 cases, most of which were stage I. Cardiac function was assessed sonographically across Quintero stages. Importantly, there was no cardiomyopathy in the donor twins. In Quintero stages I and II, 61% of recipient twins had ventricular hypertrophy, arteriovenous valve regurgitation, or quantitative abnormalities of ventricular function. Increasing prevalence of biventricular systolic dysfunction and cardiomegaly was observed with advancing Quintero stage ( Figure 5 ). This study demonstrates that pathologic changes in cardiac structure and function are present early in the evolution of TTTS. Incorporation of MPI analysis into the assessment of TTTS therefore could improve the stratification of risk and selection of cases for early treatment.
Rychik et al performed a similar assessment of recipient twin cardiac function. Fetal echocardiograms of 150 monochorionic twins were reviewed and scored with the use of a novel Cardiovascular Profile Score, which is comprised of variables such as ventricular hypertrophy, dilation, function, valve regurgitation, great artery size, and umbilical artery flow in the donor, in addition to the MPI. Blinded to Quintero stage, the Cardiovascular Profile Score was assigned to each twin set and was compared with the Quintero grade ( Figure 6 ). The Cardiovascular Profile Score correlated well the Quintero stage and also identified approximately 10% of stage I twin pairs with significant myocardial dysfunction who would have been overlooked if the Quintero criteria alone had been used.
In summary, the addition of cardiovascular parameters to the evaluation of TTTS, particularly in the early stages, appears to provide important information that may be useful to discriminate twin pairs more likely to progress to morbidity. However, the predictive ability of various cardiac indices for TTTS disease progression or adverse perinatal outcome has not been validated adequately prospectively.
Proposed alternate staging schemes
The original Quintero stage III contained a subcategory “atypical” to denote the frequent occurrence of visible bladder (stage II) with simultaneous critically abnormal Doppler findings (stage III). However, the value of having a classic stage III (absent donor bladder, abnormal Doppler findings) and an atypical category was not evaluated until recently.
Murakoshi et al assessed a small cohort (n = 31) of women with stage III TTTS, all of whom were treated with fetoscopic laser surgery and subclassified into atypical and classic. There was a significantly higher incidence of absent or reversed umbilical artery end-diastolic velocity in the donors with stage III atypical, rather than in classic, patients (83.8% vs 53.3%; P = .004). Stage III atypical cases also had a significantly higher incidence of AA anastomoses (72.9% vs 17.8%; P < .001) and donor fetal death (43.2% vs 13.3%; P = .002). This study, which indicates that abnormal Doppler findings carry greater predictive power for perinatal death than absent fetal bladder, suggests that stage II cases must be evaluated carefully for other parameters, especially umbilical and ductus venosus Doppler findings.
Rossi and D’Addario performed a metaanalysis of studies comparing Quintero stage to outcome in laser-treated TTTS pregnancies and found no significant differences in outcome by stage. Based on these findings, they proposed an alternative staging system that considers staging of donor and recipients separately: donor staging: (I) estimated fetal weight < 20th percentile, (II) reduced velocity in umbilical artery or ductus venosus, (III) fetal anemia based on middle cerebral artery peak systolic velocity Doppler finding, (IV) fetal death; recipient staging: (I) reduced velocity in umbilical artery, (II) altered MPI, (III) hydrops, (IV) death. Unfortunately this novel staging system has not been evaluated in an appropriate validation cohort or dataset.
The following possibilities, none of which have been tested in an appropriate independent cohort, include (1) for stage I, define polyhydramnios ≥8 cm at ≤20 weeks’ gestation and >10 cm at >20 weeks’ gestation, (2) combine stages I and II, (3) combine stages I and II with new substages by urinary flow rates, (4) continue Quintero staging with new substage I of AA anastomoses (yes/no), (5) continue Quintero staging with new substage I with MPI >0.4, and (6) continue Quintero staging with new substages I of AA anastomoses (yes/no) and MPI >0.4.
Conclusions for modifying the Quintero staging system (objective 1)
There are insufficient data to recommend revising or abandoning the use of the Quintero staging of TTTS. However, several other physiologic parameters that include cardiac indices and markers of systemic hemodynamic alterations appear promising for the improved prediction of TTTS disease progression and perinatal or pediatric outcome. These candidate predictive indices should be compared and validated in prospective cohorts before implementing them as standard care.
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
