valsky2010-CIR-SELECTIVO.pdf

PRENATAL DIAGNOSIS Prenat Diagn 2010; 30: 719–726. Published online 27 May 2010 in Wiley InterScience (www.interscience.wiley.com) DOI: 10.1002/pd.2536 REVIEW Selective intrauterine growth restriction in monochorionic diamniotic twin pregnancies Dan V. Valsky1,2,3 , Elisenda Eixarch1,2 , Josep Maria Martinez1 and Eduard Gratacós1,2 * 1 Department of Maternal-Fetal Medicine (Institut Clı́nic de Ginecologia, Obstetrı́cia i Neonatologia), Hospital Clinic-IDIBAPS, University of Barcelona, Barcelona, Spain 2 Centro de Investigación Biomédica en Red en Enfermedades Raras (CIBER-ER), Barcelona, Spain 3 Department of Obstetrics and Gynecology, Hadassah-Hebrew University Medical Centers, Mt Scopus, Jerusalem, Israel Selective intrauterine growth restriction (sIUGR) occurs in 10 to 15% of monochorionic (MC) twins, and it is associated with a substantial increase in perinatal mortality and morbidity. Clinical evolution is largely influenced by the existence of intertwin placental anastomoses: pregnancies with similar degrees of fetal weight discordance are associated with remarkable differences in clinical behavior and outcome. We have proposed a classification of sIUGR into three types according to umbilical artery (UA) Doppler findings (Inormal, II-absent/reverse end-diastolic flow, III-intermittent absent/reverse end-diastolic flow), which correlates with distinct clinical behavior, placental features and may assist in counseling and management. In terms of prognosis, sIUGR can roughly be divided in two groups: type I cases, with a fairly good outcome, and types II and III, with a substantial risk for a poor outcome. Management of types II and III may consist in expectant management until deterioration of the IUGR fetus is observed, with the option of cord occlusion if this occurs before viability. Alternatively, active management can be considered electively, including cord occlusion or laser coagulation. Both therapies seem to increase the chances of intact survival of the larger fetus, while they entail, or increase the chances of, intrauterine demise of the IUGR fetus. Copyright  2010 John Wiley & Sons, Ltd. KEY WORDS: selective intrauterine growth restriction; monochorionic twins; arterio-arterial anastomosis; Doppler INTRODUCTION The term selective intrauterine growth restriction (sIUGR) in monochorionic (MC) twin pregnancies is used to define cases with an estimated fetal weight (EFW) of less than the 10th percentile in one fetus. Intertwin EFW discordance above 25%, calculated as [(larger twin—smaller twin)/larger twin], is a common element accompanying this condition. Although various diagnostic criteria have been used in the literature, including EFW (Sebire et al., 1997; Gratacos et al., 2004a) or fetal weight discordance (Sebire et al., 1997; Victoria and Mora Gand Arias, 2001; Gratacos et al., 2004a; Vanderheyden et al., 2005), the above definition is widely accepted and seems to be the simplest for practical and investigational purposes. The clinical significance of MC pregnancies with both twins having an EFW of less than the 10th percentile, but without intertwin discordance, or those with intertwin discordance but a smaller fetus above the 10th percentile remains to be established. The rate of sIUGR has been reported to be 10 to 15% (Sebire et al., 1997; Lewi et al., 2008) in MC twin pregnancies. This condition represents an important contributor to perinatal mortality and morbidity in MC twins, and it is associated with a high risk of *Correspondence to: Eduard Gratacós, Department of MaternalFetal Medicine (ICGON), Hospital Clinic, Sabino de Arana 1, 08028, Barcelona, Spain. E-mail: gratacos@clinic.ub.es Copyright  2010 John Wiley & Sons, Ltd. neurological damage in both fetuses (Gratacos et al., 2004a; Ishii et al., 2009). The introduction of detailed ultrasonographic and Doppler evaluation and the development of fetoscopic techniques and placental studies have contributed to increasing our understanding, and the diagnostic and therapeutic capabilities in this condition. This review will summarize the unique aspects of the pathophysiology of sIUGR in MC twins and its implications for the diagnosis and clinical presentation. Likewise, a classification system to differentiate the main clinical forms will be discussed together with some considerations on management. UNIQUE ASPECTS OF sIUGR IN MC TWINS The main cause for the development of sIUGR in MC twins is the existence of unequal placental sharing. The relationship between unequal placental territory and birth weight discordance has recently been clearly shown in several studies (Denbow et al., 2000; Fick et al., 2006; Lewi et al., 2007; Hack et al., 2008). The presence of very eccentric or velamentous cord insertion in the smaller twin, reported to be as high as 45%, is a characteristic accompanying finding of sIUGR (Machin, 1997; Hanley et al., 2002), which has been proposed to partially contribute to this condition. However, whether it is a cause or the consequence of the extremely asymmetric distribution of placental territories remains unclear. Received: 28 December 2009 Revised: 4 March 2010 Accepted: 20 March 2010 Published online: 27 May 2010 720 D. V. VALSKY et al. Aside from placental asymmetry, the incidence and natural history of sIUGR in MC twins is different from those in singletons and dichorionic twins due to the presence of intertwin placental anastomoses (Lewi et al., 2007). Inter-fetal vascular connections, either arteriovenous (AV) (in reality a vascular aterio-venous connection but not a real anastomosis) or arterio-arterial (AA) and veno-venous (VV) (direct vascular connections and therefore real anastomoses) have a strong influence in the evolution of the sIUGR, as they interfere in the relation between placental territory discordance and birth weight discordance, which in the absence of anastomoses should be roughly linear. There seems to be a relative relation between the relative placental area and the anastomotic pattern (Lewi et al., 2007). Intertwin anastomotic area, net arterio-venous transfusion and the diameter of AA anastomoses have been shown to be correlated with the degree of intertwin placental discordance. Thus, in general, larger degrees of placental discordance are associated with more intertwin anastomoses, which results in a more pronounced dependency of the smaller twin on the circulation of its co-twin. While this may clearly be protective to the smaller twin, it may entail additional risks to both fetuses, as discussed later in this review. In addition, the above relation between placental discrepancy and inter-fetal anastomoses is not always present, and therefore it is possible to observe cases with large placental discordance but a small number of anastomoses. In clinical practice, the balance between placental territory and anastomotic pattern will finally determine the magnitude and the degree of (normally, beneficial) interference with the natural history of growth restriction. The potential resulting combinations are obviously multiple, but a simplification could be made in three major scenarios. The first two represent the opposite ends of a spectrum. At one end, there is little placental discordance and/or large arterio-venous flow interchange. At the other end is severe placental discordance with little arterio-venous interchange, which as discussed above is a rare but possible combination. The first end is associated with a benign clinical evolution. As we approach the other end, with smaller placental territories and reduced intertwin vascular connections, the likelihood for a poor outcome will increase. Indeed, in the presence of very small intertwin flow interchange, the IUGR fetus tends to behave like a dichorionic twin, where placental territory is the major determinant of IUGR. The third scenario represents a particular situation where the major distinctive feature is the presence of a large AA anastomosis. This type of vessels allows transfer of large amounts of blood. It is relatively common that these cases have larger placental area discrepancies, but in spite of this, the degree of inter-fetal weight difference is relatively lower due to shunting of large amounts of blood from the larger to the smaller twin through the large AA vessel. As a trade-off, the presence of a large AA anastomosis creates an unstable hemodynamic system and a higher risk of acute feto-fetal transfusion episodes in the event of fetal death or sustained bradycardia of the smaller twin (Gratacos et al., 2004a, 2007; Ishii et al., 2009). It is tempting to postulate that MC twins with large AA Copyright  2010 John Wiley & Sons, Ltd. vessels and very small placental territory might represent the last step in the spectrum of placental discordance in which viability of a normal fetus is preserved, with the next step being the acardiac twin with TRAP (twin reverse arterial perfusion) sequence (Figure 1). IMPLICATIONS OF THE MONOCHORIONIC PLACENTAL FEATURES IN DOPPLER FETOPLACENTAL EXAMINATION In singleton pregnancies complicated by early IUGR, the reduction in placental vascular bed occurring in placental insufficiency is normally reflected in the umbilical artery (UA) Doppler waveforms. Changes in UA pulsatility are interpreted as a surrogate parameter of increased vascular resistance in the placental bed, and such changes are consequently associated with a relatively predictable deterioration in the UA flow characterized by progressive reduction in diastolic flow and eventually absent and reverse diastolic velocities. This is accompanied from early stages by changes in middle cerebral artery (MCA) pulsatility indicating redistribution of blood flow to the brain, and in advanced stages, by increased pulsatility in venous flow parameters, which correlate with the establishment of subclinical cardiac function as a consequence of fetal hypoxia and finally acidosis (Sebire, 2003; Baschat, 2004; Turan et al., 2008). For the above-discussed reasons, the clinical interpretation of UA Doppler changes cannot be made on the same grounds as for singleton pregnancies (Hecher et al., 1994; Gaziano et al., 1998; Wee et al., 2003; Gratacos et al., 2004b). The combination of inadequate placental sharing with the presence of inter-fetal anatomoses will determine three major types of Doppler Figure 1—Monochorionic placenta in a type III sIUGR. The dotted white line indicates the vascular equator. Note the small placental area of the IUGR fetus, which is connected to the larger one through a large AA (red arrows), which connects directly the umbilical cords and allows transfer of large amounts of blood to the smaller fetus. While the ratio of the larger to smaller placental territory was 3.5 to 1, the ratio in fetal weights was to 1.75 (1750 and 1010 g, respectively). This difference illustrates the influence of the monochorionic placenta in the natural history of growth restriction. Placenta perfused with dye courtesy of Dr Maria Angeles Gomez, Buenos Aires Prenat Diagn 2010; 30: 719–726. DOI: 10.1002/pd sIUGR IN MONOCHORIONIC DIAMNIOTIC TWIN PREGNANCIES (a) (b) 721 characteristic cyclic UA flow pattern is associated with the co-existence of sIUGR, thus of significant fetal weight discordance and a large diameter of the AA anastomosis (Taylor et al., 2003; Wee et al., 2003). The classification into types was proposed because each of these Doppler patterns (Gratacos et al., 2004a) appears normally from very early stages of pregnancy, and in the vast majority of cases, it remains unchanged until delivery, and (Sebire et al., 1997) it is associated with a distinct clinical course with respect to the other types (Wee et al., 2003; Gratacos et al., 2004b; Vanderheyden et al., 2005; Ishii et al., 2009). Thus, the UA Doppler pattern observed at the time the diagnosis of sIUGR is established can be used to predict the likely clinical evolution, and this has important implications for counseling and management. CLINICAL EVOLUTION AND PERINATAL OUTCOME OF sIUGR IN MC TWINS ACCORDING TO THE TYPE OF UA DOPPLER PATTERN Type I Figure 2—Pulsed Doppler showing intermittent absent or reverse end-diastolic flow (iAREDF) in the UA of the smaller fetus (a). In most, but not all, cases, the systolic waveforms have a characteristic oscillation reflecting the influence of the transmitted waveforms on the peak velocity. The pattern of iAREDF could be more or less prominent in each case and must be actively searched for as it can be easily missed and classified as either type I or II Doppler. The sign is more easily observed at the placental cord insertion during maternal breath holding and using a low sweep speed of the spectral Doppler, which helps to identify not only the diastolic changes but also the characteristic oscillation of the systolic peak velocities. These changes are the reflection in the smaller fetus’ cord of the characteristic periodic bidirectional pattern of AA anastomosis (b), which results from the collision of two opposite systolic waveforms waveform in the IUGR twin. We have proposed to classify these patterns into types: type I with positive diastolic flow, type II with persistently absent or reverse, and type III with intermittent absent or reverse enddiastolic flow (iAREDF). The type III Doppler pattern is unique to MC twins, and it is characterized by the presence of cyclic or intermittent absence or reverse of diastolic wave, which in reality is a transmitted pattern reflecting the existence of a large placental AA anastomosis (Hecher et al., 1994; Wee et al., 2003) (Figure 2). It must be stressed that this inference does not work in the opposite direction, and therefore the presence of AA anastomoses in an MC pregnancy does not entail that iAREDF will be observed. The appearance of the Copyright  2010 John Wiley & Sons, Ltd. IUGR twins with present diastolic flow have been shown to generally have a favorable outcome. In a prospective study, a group of 39 type I twin pairs had a later gestational age at diagnosis (mean 23 weeks), later gestational age at time (mean 35.4 weeks), higher birth weight and lower intertwin discordance rate (29%) than those observed for type II or III sIUGR pregnancies (Gratacos et al., 2007). The rate of intrauterine fetal death (IUFD) was less than 3%, and the rate of parenchymal brain damage in surviving twins was 0%. These results are in agreement with another collaborative study (Ishii et al., 2009), where the outcome of 63 cases with sIUGR without signs of twin-to-twin transfusion syndrome (TTTS) was evaluated. Perinatal outcomes for type I sIUGR were, in general, much better than those observed in the other types, with an intact survival rate in both twins over 90%, an average gestational age at delivery of 36 weeks and a 4.3% rate of IUFD in the smaller twin. Placental studies have shown that each UA Doppler type in MC twins with sIUGR is roughly associated with one of the three placental patterns discussed in the preceding section. Type I Doppler is observed in pregnancies belonging to the good prognosis end of the spectrum as defined above, with a less marked degrees of intertwin placental discordance and/or a sufficient anastomotic area to support the smaller fetus. Type II Observation of AREDF in the IUGR twin is associated with early fetal deterioration in the majority of cases. Even if the average latency time from diagnosis of AREDF to delivery is much longer as compared with that reported for singleton pregnancies, 54 versus 11 days, respectively (Vanderheyden et al., 2005), either the placental insufficiency is particularly severe or the intertwin blood flow interchange cannot fully Prenat Diagn 2010; 30: 719–726. DOI: 10.1002/pd 722 D. V. VALSKY et al. compensate the placental discordance, and thus from a certain gestational age, fetal deterioration is observed. In a small proportion of fetuses, prolonged survival of the growth restricted fetus until 32 weeks or later has been observed, but this is an exception. Analysis of the placental patterns in type II pregnancies (Taylor et al., 2000) showed that these are normally associated with smaller placental territories in the IUGR twin and/or with a relatively smaller number of anastomoses. In the series published by Gratacos et al., in 30 type II patients (persistent AREDF in UA of IUGR twin) (Gratacos et al., 2007), the degree of placental discordance was high (38%) and the deterioration of the IUGR twin, as defined by abnormal venous Doppler flow or biophysical profile (BPP), was observed in 90%, with a mean gestational age at delivery of 30 weeks. It should be noted that the rate of neonatal brain damage according to ultrasound scans was 14.4% in the smaller twin. In spite of the large proportion of cases presenting with signs of fetal deterioration, there were no cases of unexpected IUFD in this group, which reflects that deterioration follows a relatively predictable evolution in most instances. Brain damage of the normally grown twin was a rare event in this series (1/30). In the study of Ishii et al. (2009), type II patients had the worst prognosis among the three study groups. An important difference of this study with respect to the previous one was that in Ishii’s series, selective feticide was not a management option due to legal constraints. Thus, cases presenting with fetal deterioration had to be managed either expectantly or by elective delivery. Accordingly, gestational age at delivery was shorter, 28 weeks, and intact survival of the smaller twin was only 37.0%, with a perinatal death rate close to 50%. The larger rate of preterm elective deliveries and IUFD had a clear impact on the outcome of the normally grown twin, which showed an intact survival of only 55%. These data are important, because they support the notion that type II cases have in reality the poorest outlook among all MC twins with sIUGR; however, this may not be apparent to clinicians if fetal cord occlusion is offered among the therapy options. Previous studies have reported the outcome of MC pregnancies with sIUGR and UA Doppler findings consistent with type II as defined in this review. In a study by Quintero et al. (2001), the outcome of 17 patients with sIUGR and absent or reverse UA Doppler managed conservatively was compared with that of 11 patients with the same Doppler changes and treated by laser coagulation of placental anastomoses. There were no differences in gestational age at delivery and the rate of at least one survival between the groups. The rate of IUFD was 41% in the conservatively treated group, with 42% of these cases complicated with concomitant death of the co-twin. The rate of severe neurological handicap was 13.6% (3/22) in newborns managed expectantly and 0% (0/12) when patients were managed by laser ablation, but the difference did not reach statistical significance. In a study by Huber et al. (2006) the natural history of MC twin pregnancies with discordant amniotic fluid volume and no signs of TTTS was analyzed. Nineteen of the pregnancies included in the study had signs consistent Copyright  2010 John Wiley & Sons, Ltd. with sIUGR in combination with AREDF in the UA, and could therefore be regarded as type II sIUGR. The overall survival rate was 60%, survival of both fetuses was observed in 47% and median gestational age at delivery was 32 weeks. Type III Type III pregnancies represent a distinct situation with respect to the other types, as clinical evolution is much less predictable. In the series published by Gratacos et al. (2007), the average fetal weight discordance was 36%, and the mean gestational age at delivery was 32 weeks. The rate of deterioration of the IUGR fetus was low (10.8%), but in spite of this apparently benign evolution, unexpected IUFD of the smaller twin occurred in 15.4% of cases, which was associated with the death of the co-twin in one-third of cases, leading to intrauterine death of the larger fetus in 6.2% of cases. The rate of neonatal brain injury in the larger twin was 19.7%, and most cases of brain injury occurred in cases with double twin survivors (10 of 12 cases). This atypical clinical evolution in pregnancies with type III sIUGR is thought to reflect the instable hemodynamic situation created by the presence of a large AA anastomosis. On the one hand, the magnitude of the intertwin blood exchange allows that the normal twin supports survival of the smaller one. This occurs at the expense of the creation of a hyperdynamic circulation in the larger twin, similar to what is found in monochorionics with an acardiac fetus or large fetal tumors, and it is reflected in the relatively common prevalence of hypertrophic cardiomyopathy-like (HCL) changes (Muñoz-Abellana et al., 2007). However, the existence of a high flow interchange through a large vessel leads to massive fetofetal transfusion episodes in the event of a sudden drop in fetal blood pressure or, more commonly, heart rate. The existence of such episodes has been documented (Gratacos et al., 2004a) (Figure 3), and it is proposed to be one of the main factors to explain the occurrence of unexpected fetal death of the IUGR and of brain injury in the normal twin, if both fetuses are born alive. In another prospective study, where the outcome of 42 MC twin pregnancies with sIUGR was compared with 32 MC uncomplicated pregnancies, Gratacos et al. (2004a) reported a significantly higher risk of IUFD (10.7 vs 0%) in pregnancies with sIUGR, mostly in the smaller twin (14.2%), and a significantly higher rate of parenchymal brain damage in the larger twin (20.5 vs 2.8% in the smaller twin). When the pregnancies with sIUGR were divided into two subgroups, depending on the UA flow of the smaller twin (with and without iAREDF), it was clearly shown that all cases of IUFD and most cases of parenchymal brain damage were associated with intermittent pattern of flow in UA. In type III cases reported in Japanese patients (Ishii et al., 2009) in accordance with previous studies, gestational age at delivery was 31 weeks, with 15.4% of smaller twin presenting with IUFD, and up to 38.5% of surviving larger twins showing brain damage at Prenat Diagn 2010; 30: 719–726. DOI: 10.1002/pd sIUGR IN MONOCHORIONIC DIAMNIOTIC TWIN PREGNANCIES 723 common to observe oscillations in the pulsatility index (PI) week by week. In most of the type I pregnancies, the sIUGR fetus will persist with a normal Doppler until advanced stages of pregnancy, allowing elective delivery at around 34 to 35 weeks. Type II Figure 3—Pulsed Doppler of the umbilical artery of the smaller twin during fetal bradycardia in a case of sIUGR type III, showing transient acute feto-fetal transfusion through the large AA anastomosis. Note the bidirectional blood flow; the retrograde systolic waveforms from the larger twin are periodically interrupted by forward systolic waveforms from the bradycardic smaller fetus. The existence of such episodes is proposed to be one of the mechanisms to explain the occurrence of unexpected fetal death of the IUGR and of brain injury in the AGA twin, if both fetuses are born alive 6 months of age. While the two studies are not directly comparable, as the time point to establish the presence of brain damage was different, the findings are consistent in pointing at an increased risk of brain injury in the normally grown twin. Concerning placental features, and as discussed above, the observation of iAREDF indicates by definition the presence of a large placental AA anastomosis (Taylor et al., 2000; Wee et al., 2003; Gratacos et al., 2004a, 2004b; Vanderheyden et al., 2005) and, normally, a much smaller placental territory in the smaller fetus. PROPOSED MANAGEMENT STRATEGY OF sIUGR IN MC TWINS Systematic follow-up of fetal growth to rule out significant EFW discordance should be a part of the routine assessment of MC twins from very early stages of pregnancy. As discussed above, both the presence of sIUGR and the type of UA flow can be determined from the early second trimester in most instances. As this pattern is not likely to change in a majority of cases, management schemes and parents’ counseling can be adjusted accordingly. Type I The evolution will be benign in most instances and therefore a policy of close follow-up to rule out evolution of UA Doppler to a different pattern, which is rare, seems reasonable. In the absence of such progression, expectant management with 1 to 2 weeks scans can be adjusted according to the magnitude of fetal weight discordance and the existence of increased pulsatility in the UA, which could be seen occasionally. It must be stressed that only the existence of AREDF has been consistently associated with poor outcome, and that it is relatively Copyright  2010 John Wiley & Sons, Ltd. The natural history of sIUGR with severe abnormalities in the UA Doppler in the form of persistent AREDF is very poor, and represents a serious challenge to the normal fetus due to the risk of IUFD of the IUGR and of the high chances of a very premature delivery. In general, the deterioration of the IUGR fetus can be anticipated to allow elective delivery but preterm birth, normally below 32 weeks, is an inevitable complication in most of these pregnancies. Expectant managed in these pregnancies will normally result in fetal hemodynamic deterioration before 32 weeks, with only rare cases progressing uneventfully to later gestational ages (Gratacos et al., 2004a, 2007; Ishii et al., 2009). Prediction of the precise timing of such deterioration in each individual case would be critical to establish the follow-up and management strategy. Prognostic factors for the timing of deterioration of type II sIUGR cases have not been formally investigated. Clinical experience, supported by biological plausibility, suggests that cases with earlier diagnosis, higher degree of discordance and more pronounced changes in the UA Doppler will probably deteriorate earlier and more rapidly. Thus, it seems logical to expect that a case diagnosed late in the second trimester with a 30% fetal weight discordance and absent diastolic wave in the UA will typically have better prognosis and deteriorate later than a case diagnosed in early second trimester with 50% discordance and reversed UA diastolic flow. While this may help some parents in deciding the management strategy, close fetal monitoring is mandatory in any event. The appearance of absent or reverse atrial flow in the ductus venosus (DV) allowed to identify severe deterioration and identify cases at risk for fetal death in prospective series (Gratacos et al., 2004a, 2007), allowing active management—that is delivery or fetal therapy—before the occurrence of intrauterine death in all reported cases. Thus, if the diagnosis of type II sIUGR is done in early stages of pregnancy, parents must be informed of the expected outcome according to published experience, and a decision should be made between expectant or active management, as cord occlusion or placental laser coagulation (see discussion on therapy options later in this review). In the cases managed expectantly, a weekly follow-up scheme is chosen, if venous Doppler is normal, with more frequent scans when PI of DV becomes elevated above the two standard deviations seems reasonable. BPP should be included in the follow-up protocol on later stages of pregnancy. When expectant management is chosen, and where late cord occlusion is legally possible, a second important decision is to define the cut-off for a decision between an elective delivery or fetal cord occlusion. For most parents and physicians, Prenat Diagn 2010; 30: 719–726. DOI: 10.1002/pd 724 D. V. VALSKY et al. a limit of 28 weeks would seem reasonable, but this might be moved to 26 or 30 weeks according to parents’ wishes. Type III For type III pregnancies, management represents a challenge. As opposed to type II, in a majority of cases, the IUGR fetus will remain stable allowing delivery at around 32 to 34 weeks, although an increased risk of sudden fetal death of the IUGR fetus and brain injury in the normally grown twin is to be expected. Again, predictive factors for a poor outcome are not well established. Clearly, an early diagnosis of sIUGR, with a high degree of discordance, prominent intermittent reverse flow in the smaller twin, and short distance between placental cord insertions are high-risk factors. When all these factors are present, the chances for a poor outcome are likely to be high and the option of elective fetal therapy might be discussed with parents. However, while these risk factors might be taken into account and help to offer rough estimates about the prognosis for each specific case, it is certainly impossible to quantify a probability of risks on the basis of the available evidence. When expectant management is decided, it should follow the same principles as discussed above for type II cases, with the important difference that in type III sIUGR, the smaller fetus will seldom present signs of deterioration in venous Doppler. In the absence of abnormal venous Doppler or BPP, elective delivery between 32 and 34 weeks of gestation is probably a reasonable option in conservatively managed cases. FETAL THERAPY FOR MC TWINS WITH sIUGR There are two relevant issues of discussion, firstly, the indication for fetal therapy and secondly, the type of procedure to be chosen. Indications for fetal therapy in sIUGR It is, and will probably remain, very difficult to establish criteria for fetal therapy in sIUGR pregnancies. Whatever the available information is, it will be unlikely to result in an optimal solution because this will always depend on three issues: (1) the risks of intrauterine fetal death and/or brain injury, (2) parents’ wishes and (3) technical considerations. For the above-discussed reasons, in most cases, only rough estimations of ‘fairly good’ or ‘rather poor’ prognosis can be made, but the chances of serious damage or death clearly exist and it is unquestionable that the presence of abnormal UA Doppler in MC twins with sIUGR, either in the form of type II or type III, is a high-risk situation. Certainly, the individual risk may vary substantially, and it is possible that any type of active therapy in utero represents more risks than benefits in certain cases. However, some parents may prefer to face this risk either to maximize the chances of having a normal baby and/or to avoid the Copyright  2010 John Wiley & Sons, Ltd. uncertainty of expectant management. Finally, technical issues may also play a role in the decision-making process, as placental location and cord insertions might render some procedures more risky or simply impossible. Cord occlusion versus placental laser coagulation Umbilical cord occlusion for selective feticide has been extensively reported as an option for complicated or discordant monochorionic twins (Ilagan et al., 2008; Rossi et al., 2009). Cord occlusion is a straightforward treatment, preventing exsanguination of the healthy fetus into the dead co-twin. The outcome of the technique has not been evaluated in prospective studies including exclusively sIUGR cases, but all published series contain such cases. Thus the chances of survival for the larger fetus range from 80 to 85%. If selective feticide is not an acceptable alternative for parents, or in countries where it is not a legal option (Ishii et al., 2009; Rossi et al., 2009), laser coagulation of placental anastomoses may be offered. In a small clinical series (Gratacos et al., 2008), 18 type III sIUGR cases were treated with laser. Laser coagulation was reported to be technically difficult and not feasible in two cases. Placental laser dichorionization resulted in a 66.7% (12/18) rate of IUFD of the smaller twin after the procedure, a significantly larger proportion than the 19.4% (6/31) of cases managed conservatively during the same study period. However, the risk of concomitant death of the larger twin was significantly lower [0% (0/12) vs 50% (3/6), p = 0.02]. The rate of periventricular leucomalacia in the larger twin in the laser group was lower than in cases managed conservatively [5.9% (1/17) vs 14.3% (4/28)]. However, the difference did not reach statistical significance (p = 0.63). As previously mentioned, laser therapy was evaluated as a treatment option for type II sIUGR in a study by Quintero et al. (2001). In this report, both fetuses survived in 4/11 treated pregnancies. Five out of eleven (45%) IUGR fetuses died after selective laser photocoagulation of communicating vessels. As discussed above, laser treatment in this group of patients was associated with a high postoperative fetal death rate in comparison with conservatively managed group. No significant protective effect in the rate of severe neurological impairment was shown after laser procedure. One major drawback of laser is that the technique may not be feasible in cases where an anterior placenta is combined with large AA vessels and placental cord insertions, which are close to each other (Gratacos et al., 2008). Thus, as discussed above, technical issues may eventually influence the decision, irrespective of the severity of the case and parents’ wishes. Therefore, even in cases of extreme severity and obvious fetal deterioration, laser may not be an option, and if a cord occlusion is not acceptable for any of the above reasons, the case will necessarily have to be managed expectantly. In summary, although there are no large series to counsel parents with accurate figures, in general, both Prenat Diagn 2010; 30: 719–726. DOI: 10.1002/pd 725 sIUGR IN MONOCHORIONIC DIAMNIOTIC TWIN PREGNANCIES cord occlusion and placental laser coagulation seem to increase the chances of intact survival of the larger fetus, while they increase the risk of intrauterine demise for the IUGR fetus. LATE ONSET sIUGR IN MC TWINS This review has focused in early sIUGR, characterized by an early onset and diagnosed around 20 weeks of gestational age and which seems to be the most common form of sIUGR presenting in MC twins. Recent clinical series have described the outcome of late onset sIUGR, diagnosed after 26 weeks (Lewi et al., 2008). Lewi et al. (2008) compared a group of 13 MC discordant twins with late onset sIUGR with a group of early onset discordance and uncomplicated MC pregnancies. Late sIUGR was characterized by an almost concordant growth until late second trimester, with sIUGR developing progressively from then onward. Intertwin fetal weight reached on average 30%, and mean gestational age at delivery was 35 weeks. UA Doppler was normal in all cases and there was one case of IUFD (8%) in this series. The rate of intertwin hemoglobin discordance, meeting the criteria for twin anemia-polycytemia sequence, was 38%. Placental studies revealed a relatively low larger to smaller placental ratio of 1.59, as compared with 2.55 in early onset IUGR pregnancies and 1.38 in controls, and a significantly lower total anastomotic diameter of 5.83, in comparison with early onset cases and controls (13.97 and 8.86, respectively), with a few small, mostly unidirectional, anastomoses. The pathogenesis of such late discordance is thought to be the establishment of a chronic unidirectional intertwin blood transfusion toward the larger twin through small AV anastomoses (Lewi et al., 2008). The practical implication of such findings is in the use of measurement of peak systolic velocity of MCA as a part of routine evaluation in cases with late appearance of sIUGR for the diagnosis of fetal anemia-polycytemia sequence in order to plan an optimal timing for elective delivery and appropriate postpartum neonatal management. CONCLUSION There is growing evidence supporting a connection between a particular placental architecture with different clinical types of discordant growth in monochorionic twins. While unequal placental sharing appears to be the main factor in the etiology of early onset sIUGR, intertwin transfusion imbalance seems to be involved in late discordant growth. A classification of early sIUGR into three types according to UA Doppler findings seems to correlate with specific patterns of placental anatomy, clinical evolution, and fetal outcome. This classification may be useful for planning the management strategy. It should be emphasized that the evidence base for this field is still relatively small and mainly based on observational studies. Unfortunately, considering the small number of cases, the variability in clinical presentation Copyright  2010 John Wiley & Sons, Ltd. and the parents’ preferences in relation with the management options, a randomized controlled trial of intervention is unlikely to be undertaken. Ideally, an international registry could help to establish large numbers of cases that would help advancing in the knowledge of this field. ACKNOWLEDGEMENTS E.E. is supported by a Rio Hortega research fellowship from the Spanish Fondo de Investigaciones Sanitarias. REFERENCES Baschat AA, Hecher K. 2004. Fetal growth restriction due to placental disease. Semin Perinatol Feb;28(1): 67–80. Denbow ML, Cox P, Taylor M, Hammal DM, Fisk NM. 2000. Placental angioarchitecture in monochorionic twin pregnancies: relationship to fetal growth, fetofetal transfusion syndrome, and pregnancy outcome. Am J Obstet Gynecol Feb;182(2): 417–426. Fick AL, Feldstein VA, Norton ME, et al. 2006. Unequal placental sharing and birth weight discordance in monochorionic diamniotic twins. Am J Obstet Gynecol 195(1): 178–183. Gaziano E, Gaziano C, Brandt D. 1998. Doppler velocimetry determined redistribution of fetal blood flow: correlation with growth restriction in diamniotic monochorionic and dizygotic twins. Am J Obstet Gynecol 178(6): 1359–1367. Gratacos E, Antolin E, Lewi L, et al. 2008. Monochorionic twins with selective intrauterine growth restriction and intermittent absent or reversed end-diastolic flow (Type III): feasibility and perinatal outcome of fetoscopic placental laser coagulation. Ultrasound Obstet Gynecol 31(6): 669–675. Gratacos E, Carreras E, Becker J, et al. 2004a. Prevalence of neurological damage in monochorionic twins with selective intrauterine growth restriction and intermittent absent or reversed end-diastolic umbilical artery flow. Ultrasound Obstet Gynecol 24(2): 159–163. Gratacos E, Lewi L, Carreras E, et al. 2004b. Incidence and characteristics of umbilical artery intermittent absent and/or reversed end-diastolic flow in complicated and uncomplicated monochorionic twin pregnancies. Ultrasound Obstet Gynecol 23(5): 456–460. Gratacos E, Lewi L, Munoz B, et al. 2007. A classification system for selective intrauterine growth restriction in monochorionic pregnancies according to umbilical artery Doppler flow in the smaller twin. Ultrasound Obstet Gynecol 30(1): 28–34. Hack KE, Nikkels PG, Koopman-Esseboom C, et al. 2008. Placental characteristics of monochorionic diamniotic twin pregnancies in relation to perinatal outcome. Placenta Nov;29(11): 976–981. Hanley ML, Ananth CV, Shen-Schwarz S, Smulian JC, Lai YL, Vintzileos AM. 2002. Placental cord insertion and birth weight discordancy in twin gestations. Obstet Gynecol Mar;99(3): 477–482. Hecher K, Jauniaux E, Campbell S, et al. 1994. Artery-to-artery anastomosis in monochorionic twins. Am J Obstet Gynecol 171(2): 570–572. Huber A, Diehl W, Zikulnig L, Bregenzer T, Hackelöer BJ, Hecher K. 2006. Perinatal outcome in monochorionic twin pregnancies complicated by amniotic fluid discordance without severe twintwin transfusion syndrome. Ultrasound Obstet Gynecol Jan;27(1): 48–52. Ilagan JG, Wilson RD, Bebbington M, et al. 2008. Pregnancy outcomes following bipolar umbilical cord cauterization for selective termination in complicated monochorionic multiple gestations. Fetal Diagn Ther 23: 153–158. Ishii K, Murakoshi T, Takahashi Y, et al. 2009. Perinatal outcome of monochorionic twins with selective intrauterine growth restriction and different types of umbilical artery Doppler under expectant MANAGEMENT. Fetal Diagn Ther 26(3): 157–161. Lewi L, Cannie M, Blickstein I, et al. 2007. Placental sharing, birthweight discordance, and vascular anastomoses in monochorionic Prenat Diagn 2010; 30: 719–726. DOI: 10.1002/pd 726 D. V. VALSKY et al. diamniotic twin placentas. Am J Obstet Gynecol Dec;197(6): 587.e1–587.e18. Lewi L, Gucciardo L, Huber A, et al. 2008. Clinical outcome and placental characteristics of monochorionic diamniotic twin pairs with early- and late-onset discordant growth. Am J Obstet Gynecol Nov;199(5): 511.e1–511.e17. Lopriore E, Deprest J, Slaghekke F, et al. 2008. Placental characteristics in monochorionic twins with and without twin anemiapolycythemia sequence. Obstet Gynecol Oct;112(4): 753–758. Machin GA. 1997. Velamentous cord insertion in monochorionic twin gestation. An added risk factor. J Reprod Med Dec;42(12): 785–789. Muñoz-Abellana B, Hernandez-Andrade E, Figueroa-Diesel H, et al. 2007. Hypertrophic cardiomyopathy-like changes in monochorionic twin pregnancies with selective intrauterine growth restriction and intermittent absent/reversed end-diastolic flow in the umbilical artery. Ultrasound Obstet Gynecol Dec; 30(7): 977–982. Rossi AC, D’Addario V. 2009. Umbilical cord occlusion for selective feticide in complicated monochorionic twins: a systematic review of literature. Am J Obstet Gynecol 200: 123–129. Sebire NJ. 2003. Umbilical artery Doppler revisited: pathophysiology of changes in intrauterine growth restriction revealed. Ultrasound Obstet Gynecol May;21(5): 419–422. Sebire NJ, Snijders RJ, Hughes K, et al. 1997. The hidden mortality of monochorionic twin pregnancies. Br J Obstet Gynaecol 104(10): 1203–1207. Copyright  2010 John Wiley & Sons, Ltd. Quintero RA, Bornick PW, Morales WJ, Allen MH. 2001. Selective photocoagulation of communicating vessels in the treatment of monochorionic twins with selective growth retardation. Am J Obstet Gynecol 185: 689–696. Taylor MJ, Denbow ML, Tanawattanacharoen S, Gannon C, Cox PM, Fisk NM. 2000. Doppler detection of arterio-arterial anastomoses in monochorionic twins: feasibility and clinical application. Hum Reprod Jul;15(7): 1632–1636. Turan OM, Turan S, Gungor S, et al. 2008. Progression of Doppler abnormalities in intrauterine growth restriction. Ultrasound Obstet Gynecol Aug;32(2): 160–167. Victoria A, Mora Gand Arias F. 2001. Perinatal outcome, placental pathology, and severity of discordance in monochorionic and dichorionic twins. Obstet Gynecol 97(2): 310–315. Vanderheyden TM, Fichera A, Pasquini L, et al. 2005. Increased latency of absent end-diastolic flow in the umbilical artery of monochorionic twin fetuses. Ultrasound Obstet Gynecol 26(1): 44–49. Wee LY, Taylor MJ, Vanderheyden T, Talbert D, Fisk NM. 2003. Transmitted arterio-arterial anastomosis waveforms causing cyclically intermittent absent/reversed end-diastolic umbilical artery flow in monochorionic twins. Placenta Aug;24(7): 772–778. Prenat Diagn 2010; 30: 719–726. DOI: 10.1002/pd