Numerous treatments for twin to twin transfusion syndrome (TTTS) have been proposed including selective reduction, cord coagulation, sectio parva, placental blood letting, maternal digitalis therapy, maternal indocin therapy, serial amnioreduction, microseptostomy of the intertwin membrane, and nonselective or selective fetoscopic laser photocoagulation.
For decades in the United States, serial amnioreduction has been the most widely accepted therapy for TTTS, but in recent years selective fetoscopic laser photocoagulation has become more accepted and in some centers is the only treatments offered.
Amnioreduction was first employed for maternal comfort and as a means to control polyhydramnios in the hope of prolonging the pregnancy until the risks of extreme prematurity were lessened.
In addition, amnioreduction improves uteroplacental blood flow, likely by reduction of pressure from the amniotic fluid.
In uncontrolled series, amnioreduction improves survival compared to the natural history of untreated Twin-Twin Transfusion Syndrome / TTTS. Moise, in a review of 26 reports dating from the 1930's of 252 fetuses, found an overall survival of 49 percent (24).
The survival in more recent series, with more consistently aggressive serial amnioreduction to reduce amniotic fluid volume to normal, have ranged widely from as low as 37 percent to as high as 83 percent (25, 26, 27, 28, 29, 30). However, these retrospective series are comprised of small numbers of patients from a range of gestational ages, as well as from a spectrum of severity of TTTS. Severity of TTTS and gestational age at diagnosis may have a profound impact on the observed mortality with any treatment strategy. Generally, the earlier in gestation TTTS presents, the worse the prognosis.
Mari et al. found that patients presenting with advanced TTTS prior to 22 weeks' gestation and absent end diastolic flow in the recipient umbilical artery had a survival of both twins with aggressive amnioreduction of only 13 percent and, with absent end diastolic flow in the donor umbilical artery, survival was 33 percent (16).
The paradoxical resolution of oligohydramnios after only a single amnioreduction was first suggested by Saade et. al. to be due to inadvertent puncture of the intertwin membrane (31).
Intertwin septostomy was specifically proposed as a treatment for TTTS to restore amniotic fluid dynamics without the need for repeated amnioreduction. One objection to this approach is the possibility it would result in a large septostomy, creating an essentially monoamniotic sac with the attendant risk of cord entanglement. In a small multicenter series of 12 patients, Saade et al. , reported an 81 percent survival with microseptostomy (32). However, not only was this series small and uncontrolled, there was no report of neurologic or cardiac morbidity. In a direct comparison, albeit a small retrospective single institution series, of serial amnioreduction versus microseptostomy, Johnson et al. observed no survival advantage with either therapy (33). This was confirmed by Saade et al. who reported the results of a multicenter prospective randomized clinical trial comparing amnioreduction to septostomy. The survival in each arm of the study was 65 percent (34) consistent with the notion that the effect of amnioreduction may be inadvertent septostomy. These studies, however, cannot prove this theory. Ultrasound guided needle technique of creating an intertwin septostomy has been abandoned, however, because of the risk of creating a monoamniotic sac and resultant cord entanglement and risk of dual fetal demise.
Fetoscopic Laser Photocoagulation
The first treatment for TTTS that attempted to treat the anatomic basis for the syndrome was reported by DeLia et al. (35, 36) who described fetoscopic laser photocoagulation of vessels crossing the intertwin membrane.
At least in theory, this treatment option should be superior since it not only arrests shunting of blood from the donor to recipient, but also halts transfer of potential vasoactive mediators. In his first small series, DeLia reported a survival of 53 percent in 26 patients (36). While survival was not significantly better than previous reports with serial amnioreduction, the "neurologic outcome" in 96 percent of survivors was "normal" as assessed by head ultrasounds. Other groups from Europe have reported similar survival with non-selective laser photocoagulation. Ville et al, reported 53 percent survival with a fetoscopic laser technique which was better than the survival observed with historical controls at the same center with serial amnioreduction (37 percent) (37). They also observed a lower incidence of abnormalities dictated by neonatal head ultrasound compared to historical controls.
The non-selective fetoscopic laser technique photocoagulates all vessels crossing the intertwin membrane. This approach may be problematic, as the intertwin membrane often bears no relationship to the vascular equator of the placenta.
This non-selective laser photocoagulation of all vessels crossing the intertwin membrane may sacrifice vessels not responsible for the TTTS, resulting in a higher death rate of the donor twin from acute placental insufficiency (38). More recently, a selective approach to fetoscopic laser photocoagulation in TTTS has been described by Quintero et al. (38).
Unlike the non-selective coagulation technique initially described by DeLia, the selective technique does not photocoagulate every vessel crossing the intertwin membrane. Only direct, arterial-arterial and veno-venous connections are photocoagulated along with any unpaired artery going to a cotyledon with the corresponding vein (and vice versa) going to the opposite umbilical cord.
Vessels on the chorionic plate can be differentiated endoscopically because arteries usually cross over veins and are darker in color due to lower oxygen saturation.
In a non-randomized comparison of patients treated by serial amnioreduction at one center and selective laser photocoagulation at another, the overall survival was not statistically significantly different (61 percent for laser vs 51 percent for serial amnioreduction) (39). However, the survival of at least one twin with laser photocoagulation was 79 percent, while survival of at least one twin with serial amnioreduction was only 60 percent (P<0. 05) (39).
The Eurofoetus trial conducted by Senat et al. (40) was the first prospective randomized trial that compares the efficacy and safety of treatment of TTTS with laser therapy versus serial amnioreduction.
Women presenting between 15 and 26 weeks gestation with polyhydramnios in the recipient twin and oligohydramnios in the donor twin were allowed to participate.
Fifty-two percent of patients were stage I or II, 47 percent were stage III and 1 percent were stage IV. Enrollment was halted after a planned interim analysis revealed a significantly higher likelihood of survival of at least one twin to 28 days of age (76 percent versus 56 percent, P = 0. 009) and to six months of age (76 percent versus 51 percent, P = 0. 002) in the laser group compared to the amnioreduction group.
More infants were alive without neurologic abnormalities detected on neuroimaging studies in the laser group as well (52 percent versus 31 percent, P = 0. 003).
The overall survival in the laser arm was 57 percent. This is consistent with previous reports of non-selective fetoscopic laser (53 percent) (36, 37). This is significantly lower, however, than the survival reported with selective fetoscopic laser (64 to 68 percent) (41, 42). Of particular concern is the poor survival which was observed in the amnioreduction arm.
The overall survival was only 39 percent, which is significantly lower than previously reported (60 to 65 percent) (16, 34). Antenatal, peripartum and neonatal care was provided by the referring hospital and lack of standardization may explain some of these differences (43).
The decreased survival in the amnioreduction group may reflect the higher pregnancy termination rate in the amnioreduction group (16 percent versus 0 percent in the laser group). The terminations were requested after the diagnosis of severe fetal complications It would be instructive to know whether these women were offered cord coagulation as a means of rescuing one baby (43).
Reliable assessment of neurologic outcome is critical when assessing efficacy of treatment for TTTS. While there was a lower rate of abnormality on neurologic imaging in the laser group (7 percent versus 17 percent), there was no long-term neurodevelopmental assessment.
Close Proximity Cotyledons
While it makes sense to use a less invasive approach to treatment if the options give similar results, it would be prudent to move promptly to laser therapy if in rigorous studies can prove laser has better short term and long term outcomes in the setting of advanced disease.
One potential limitation to the success of laser treatment is the presence of deep vascular AV anastomoses that cannot be identified endoscopically. In one study, vascular casts of 8 of 15 placentae (53 percent) demonstrated potentially significant atypical AV anastomoses such that two apparently normal cotyledons were actually communicating below the chorionic surface (9). A second type of atypical AV anastomoses was noted in 11 of the 15 placentae (73 percent) in which shared cotyledons arise within larger apparently normal cotyledons.
We are now able to identify these shared cotyledons at the time of the fetoscopic laser procedure. Paired vessels supplying cotyledons within one fetoscopic visual field should be treated as "close proximity cotyledons". When left intact, TTTS presents, and when subsequently treated, TTTS resolves. In cases in which there is very unequal placental sharing, the twin with the smaller placental share can be allowed to keep the cotyledon and the vessels to the other twin are taken.
Adjunctive Medical Treatment of TTTS
Crombleholme et al. (2010) conducted a study to evaluate the effect of maternal nifedipine on fetal survival when started 24-48 hours before selective fetoscopic laser photocoagulation (SFLP). The results of the study are the first to demonstrate that the outcome of SFLP treatment of TTTS can be influenced by an adjunctive medical therapy with the administration of maternal nifedipine. The mechanism by which the patients benefit is unclear, but the beneficial effect of nifedipine treatment appears to be specific to recipient twins, because no significant effect is observed on donor-twin survival. The specific beneficial effect of nifedipine on recipient survival that was observed in this case control study is all the more remarkable when one considers that cases of significantly more advanced-stage TTTS were included in the nifedipine-treated group (61.0% for patients with stage IIIC and IV compared with 29.6% for control subjects; P < .001). The reason for this is that there were fewer late stage (IIIC and IV) control cases available for gestational age and stage matching. We elected to use earlier-stage case controls (ie, stage IV nifedipine-treated patients matched with a control subject of the same gestational age but at an earlier stage, for example, stage IIIA or III), which by intent biased the study against the nifedipine-treatment group. Because of this matching, the overall comparison between the nifedipine cases and control subjects across stages is balanced in terms of gestational age distribution but unfavorable to the cases in terms of TTTS stage distribution. On the contrary, comparisons in each stage are not balanced in gestational age distribution and nifedipine effect may be confounded. This may account, in part, for the lack of survival advantage that was observed in patients with stages IIIC and IV. It is also possible that the severity of the TTTS cardiomyopathy in stages IIIC and IV are too advanced to respond to the brief 24-48 hours of preoperative treatment with nifedipine to affect acute fetal survival.
There is a benefit to overall fetal survival; 83% of nifedipine-treated fetuses survived compared with 75% of the control fetuses. This benefit to overall fetal survival appears to be due to improved recipient survival in mild-to-moderate TTTS cardiomyopathy in stages IIIA (100% vs 81%) and IIIB (93% vs 71%). There is also a benefit to acute survival of recipients with nifedipine treatment (92%), compared with control subjects (83%), with little change in survival to birth, which suggests that the beneficial effect is a result of better perioperative survival with nifedipine treatment. The results of this pilot study provide proof of the concept that adjunctive medical therapy, as with nifedipine, may improve survival in cases of TTTS that are undergoing SFLP.
The comparison of pre- and postoperative fetal echocardiograms, however, could not demonstrate a significant improvement in Tei myocardial performance indices for either the LVs or RVs with nifedipine treatment. This is not surprising for several reasons. First, our group has already shown that SFLP, by itself, improves fetal echocardiographic parameters and that an additional benefit of nifedipine would likely be small and require large numbers of subjects to demonstrate a benefit. Second, patients were treated for a brief period preoperatively and reassessed only 3-5 days after SFLP, which may be insufficient time to show differences in cardiac recovery between nifedipine-treated and control fetuses. Third, only acute survivors are assessed by postoperative fetal echocardiograms; sicker patients who died after SFLP fall out of this analysis. Because we demonstrated an acute survival advantage for nifedipine-treated cases, it is possible that the perioperative deaths in the control group eliminated patients that would have had worse after SFLP echocardiographic findings. Last, improvements in the Tei myocardial performance indices are small, and there is a significant variability both within and between patients, which could require large numbers of patients to demonstrate an improvement, if present, because of variability in the data. The mechanism by which nifedipine confers this survival benefit is uncertain but may make recipient twins with TTTS cardiomyopathy better able to tolerate the hemodynamic stress, altered placental resistance, and/or increased afterload that may follow SFLP.
Echocardiographic data from our experience with severe TTTS suggest that TTTS results in hypertensive changes in the recipient twin that resolve with the arrest of the syndrome by SFLP. 12-14 Recipient fetal RV systolic pressures that were measured with echocardiography from valvular regurgitant jets consistently showed pressures in excess of 50-75mm Hg. 7; 44. In mid-gestation fetal hearts, this is an extraordinarily high pressure considering that normal fetal RV pressures are <25 mm Hg. 41. This finding often precedes advanced signs of TTTS, such as reversal of blood flow in the a wave of the ductus venosus, umbilical artery Doppler velocimetry changes, cardiac decompensation, or hydrops. We hypothesize that differential blood flow shifts in the monochorionic placenta results in high levels of placenta-derived vasoactive mediators that are shunted preferentially to the recipient twin that induce hypertensive cardiomyopathy. Interruption of these vascular communications by SFLP eliminates the hypertensive stress in the recipient twin that allows recovery of the recipient cardiac function, resolution of cardiac hypertrophy, valvular insufficiency, and even hydropic changes. 7; 12 In twins with TTTS who are treated by SFLP at our center, we have observed with echocardiography the postoperative resolution of these hypertensive changes in surviving recipient twins. 12 Consistent with the hypothesis that vasoactive mediators play an important role in TTTS, Bajoria et al l5 reported that endothelin-1, a potent vasoconstrictor, is 2.5-fold higher in the serum of recipient twins than in donor twins. Moreover, plasma endothelin-1 levels were significantly higher in the recipient twins with hydrops than in those with mild or no hydrops. 15 Autopsy results in recipient twins with TTTS that differentiated nephrosclerosis from hypertension are also consistent with this hypothesis. 42 Identifying vasoactive mediators that are responsible for TTTS would address a large gap in knowledge and potentially suggest new treatment approaches.
Nifedipine is a calcium channel blocker that antagonizes voltage-dependent L-type channels that result in vascular and extravascular smooth muscle relaxation. 26 It has a high oral absorption rate with low bioavailability (30-60%) because of substantial first-pass effect in the intestinal wall and the liver, whereby it is metabolized to less active metabolites by the oxidative pathway, mainly by the cytochrome P-450.43 It is highly protein bound (90-95%), and its metabolites are excreted mainly by the kidneys (70-80%). 26-29,44'45 Because of the aforementioned characteristics, nifedipine has a short half-life with large interindividual variability in its kinetics with the highly variable plasma concentrations after oral administration. 44 Peak serum concentrations are lower and the terminal elimination half-life (mean, 1.3 ±0.5 hours) is shorter during the third trimester, compared with levels in nonpregnant patients. 31 In addition, nifedipine is detected in samples of fetal cord blood and amniotic fluid at concentrations approximately 93% and 53% of those of simultaneously collected maternal vein samples, respectively. 31
This study was an observational case control cohort with treated and control subjects who were matched for gestational age and Cincinnati stage. The nifedipine-treated and control cohorts are from 2 different time periods during which both volume and experience with SFLP at our center increased. A possible alternative explanation for the survival benefit that was observed in recipients could be improvement in the technique over this timeframe. However, in both time periods, the same staging system, mapping protocol, and operative technique were used. It is important to note that members of our team with little experience with SFLP began operating during the time period that the nifedipine-treated cases were obtained argues against the improved recipient survival being due to improved surgical technique or experience. The retrospective nature of the study design, however, limits our ability to prove definitively that nifedipine is directly responsible for the survival benefit that was observed in the nifedipine-treated patients. A prospective randomized placebo-controlled trial will be necessary to prove definitively the beneficial effects of nifedipine on recipient cardiac function and survival in TTTS.
The apparent beneficial effect of nifedipine on survival suggests that, although the underlying pathophysiologic condition of TTTS still is understood incompletely, there is the potential for targeted medical therapy to improve outcomes. The underlying cause of TTTS remains unknown, and the positive findings in this study support greater efforts at defining the cause of TTTS to develop targeted medical therapies that might eliminate the need for invasive surgical interventions. (Crombleholme et al., 2010)
Fetoscopic Cord Coagulation / Radiofrequency Ablation
Some centers have taken the view that the most definitive approach to treating TTTS is selective reduction using fetoscopic cord ligation or coagulation. The rationale for this approach is that cord occlusion and sacrifice of one twin arrests the syndrome, prolongs the gestation, and maximizes the outcome for one twin. We have reserved this approach for instances where advanced TTTS cardiomyopathy has irretrievably compromised the recipient twin with no hope for salvage. In such cases, due to unequal sharing between the donor and recipient, the selective fetoscopic laser procedure may result in death of the donor twin from acute placental insufficiency within hours of the procedure and a recipient twin that dies from progressive TTTS cardiomyopathy. In this situation, fetoscopic cord coagulation may be the best option available.
Cord coagulation preserves the vascular communications between the donor twin and the placenta in the recipient twin's domain. In 16 of 17 such cases we have observed rebound fetal growth, restoration of amniotic fluid volume, and delivery of neurologically intact donor twin at a mean gestational age of 34 weeks. One survivor had a grade I intraventricular hemorrhage but is otherwise doing well.
SFLP vs. RFA for Stage IV TTTS
While selective reduction by RFA in stage IV TTTS may be an option, direct comparison of SFLP and RFA would suggest otherwise except in very specific cases. In a series of 50 patients with stage IV TTTS, in which 12 were treated with RFA and 38 were treated by SFLP, the overall fetal survival in the RFA treated cases was 50%. In contrast, those treated by SFLP had an overall fetal survival of 85%, with 100% having at least one survive and 70% having both twins survive. The recipient twin was invariably the hydropic twin due to TTTS cardiomyopathy. The recipient twin had on overall survival of 78% (donor survival was 93%). There were specific fetal echocardiographic parameters which predicted poor recipient survival. If the recipient twin had both pulmonic atresia and severe LV dysfunction the survival was only 20%. In the absence of pulmonic atresia and LV dysfunction survival of the recipient twin was 100%. If only one of these findings was present recipient survival was 88%.
We now reserve RFA in stage IV to highly select cases as outcomes with SFLP are so good.
Our approach has been to offer sequential therapy to patients with early TTTS stages I, II or III in the absence of TTTS cardiomyopathy tailored to the needs of a given set of twins based on gestational age at presentation and evidence or progression of hemodynamic compromise based on Doppler velocimetry and echocardiographic changes.
In this approach, only those cases, in which less invasive approaches have failed, are offered the more invasive fetoscopic treatments. In patients who present later than 24 weeks gestation, we have favored amnioreduction or microseptostomy based on the more favorable prognosis in these patients.
In patients presenting prior to 24 weeks without advanced cardiac changes in the recipient, we have leaned toward observation (stage I or II) or a trial of amnioreduction. Patients presenting earlier in gestation tend to develop signs of hemodynamic progression of TTTS more rapidly.
For this reason, all pregnancies undergo close serial sonographic and echocardiographic surveillance for progression in cardiac and hemodynamic changes which would be an indication for selective fetoscopic laser surgery.