Placenta

Placenta. Twin–Twin Transfusion Syndrome

The normal fetoplacental circulation consists of two arteries and one vein, which divide progressively in the chorionic plate (fetal surface of the placenta), and irrigate each cotyledon in a separate way. This means that normal vascularization of the placentas may be clearly recognized by the naked eye, until it reaches the corresponding cotyledon. 

All monochorionic placentas have vascular communications between the cords,5 which was demonstrated in 278 placentas from twin pregnancies. Monochorionic placentas always have vascular communications, in comparison with dichorionic fused placentas. Despite this finding, TTTS only occurs in 15% of monochorionic pregnancies. The pattern of the anastomoses is therefore determinant in the occurrence of TTTS. 

The anastomoses are classified into superficial arterioarterial (AA), superficial venovenous (VV), and deep arteriovenous (AV). Superficial anastomoses are direct vascular communications between vessels that come from each cord insertion. They may connect arteries (AA) or veins (VV) directly, and are called superficial because they are visible in the surface. Their main particularity is that they can compensate for higher volumes of blood in a rapid manner, in both directions, depending on arterial system or venous system blood pressure differences between the fetuses. Superficial arteriovenous anastomoses do not exist because they would produce a rapid exanguination of one fetus into the other. On the other hand, deep anastomoses are not real direct vascular communications, but cotyledons that are irrigated by a chorionic artery of one fetus that is drained by a chorionic vein of the other. They function in a much slower manner, and they are unidirectional (Figures 4.2, 4.3, 4.4, 4.5, and 4.6). 

The vascular anatomy of monochorionic placentas was first described at the beginning of the century.6 It is only in the 1990s that larger series were published to review the evidence of its role in TTTS. 

The study of placentas require a standardized technique and a clear clinical diagnosis, with inclusion and exclusion criteria, to avoid selection bias. 

The methods for ex-vivo evaluations described in the literature are colored ink injection, milk injection, air injection, ex-vivo angiography, and direct visualization of the placenta. In most techniques, veins and arteries catheterization is necessary, in order to identify the communicating vessels more easily and to be able to observe clear images of the vessels. 

In our experience, direct visualization is an excellent technique, with a good quality of vessel identification. The benefit of colored ink injection is that it can be easily photographed, so that post-hoc evaluations can be performed. There are no studies that compare direct visualization, air injection, or ink injection in terms of sensitivity. Injection studies have the problem of the dye viscosity and vascular thrombosis. 

This affects the perfusion of terminal circulation, and may give false-negative results. Normally, all vessels well seen by direct visualization will be well seen in ink injection studies. Rarely, a vessel not seen by the naked eye can be discovered by an ink injection. Air injection may aid in the identification of patent anastomoses after laser therapy. Finally, the death of one twin will produce an obliteration of the arteries and veins of the dead twin, making difficulty for an injection study later after delivery. 

The best cases for evaluation are those delivered before laser treatment of those with doublefetal demise in which delivery was induced soon after. The differences in etiology, natural history, and treatment options are still debatable, and are related to the differences in the studies published.7 The following studies correlate the placental findings with clinical data. A study of monochorionic placentas by Machin et al8 examined 69 placentas consecutively from monochorionic twin pregnancies that presented in their center. They studied the presence of vascular anastomoses, placental sharing, and their correlation with perinatal mortality, growth discordance, gestational age at delivery, and polyhydramnios. They provided a good classification of the 13 patterns of anastomoses. In their study, the worst clinical outcomes were the cases of unequal placental sharing, with the presence of deep AV anastomoses and paucity of superficial anastomoses (60% of TTTS within this group). They had 23% of placentas without anastomoses. 

Bajoria et al published an interesting work of 10 placentas from TTTS and 10 from normal monochorionic pregnancies.9 Exclusion criteria were the death of one or both twins and laser therapy, although the number of excluded placentas is not mentioned. They described a technique with room temperature and humidity control, right after delivery with a perfusion protocol of dextran, heparin, and procaine as vasodilator. Pressure, time intervals, and pH were controlled in the perfusion. A minimum of 25 cycles of 15 minutes each were performed for a blood-free outflow. Later, a dye injection procedure was described to ascertain the anastomoses pattern of the placenta. They found globally less anastomoses in the TTTS group (median 1 [0–2] vs 6 [4–8], p <0.001, Mann–Whitney). They found that TTTS placentas had fewer AA anastomoses (median 0 vs 2, p <0.0001), less VV anastomoses (median 0 vs 2, p <0.0001), and less deep AV anastomoses (median 1 vs 2.5, p <0.001). TTTS placentas had the particularity that they usually had one anastomosis, except in one case that had two anastomoses, in comparison to non- TTTS that presented with multiple anastomoses. The type of anastomosis was also different: whereas TTTS placentas had deep AV in 80% of the cases, non-TTTS presented them in only 36% of the cases. The deep AV in TTTS cases were always donor to recipient, in comparison to non-TTTS cases. All control cases had both AA and VV anastomoses. 

The authors conclude that is that TTTS placentas have generally one anastomosis, and that the anastomoses are deep AV without superficial ones. In controls, multiple anastomoses were present, with deep AV and superficial ones. Denbow et al10 studied 71 placentas from consecutive monochorionic–diamniotic pregnancies using a simpler technique: 82 patients were initially recruited, but exclusion criteria were applied for TRAP sequence (2), selective feticide (2), placental destruction at delivery (2), conjoined twins (2), and lost on follow-up (3). Four monoamniotic cases were included. This is the only study blinded to the perinatal outcome in which the ink injection was performed by a senior perinatal pathologist. The article explains clearly that in three cases only a limited assessment was possible, because of autolysis after fetal demise or damage at delivery. They found that TTTS placentas have fewer AA anastomoses (median 0 [0–1] vs 1 [0–1], p <0.0001), with similar frequency of AV and VV anastomoses. The highest rate of TTTS was found in placentas with one AV and no AA anastomoses (78%). All TTTS placentas had deep AV anastomoses, in comparison to 84% of non-TTTS placentas. The direction of AV anastomoses is not described in the study. TTTS cases with death of one twin were no different than those in which both survived. The authors do not describe difficulties in catheterizing the umbilical cord in the dead twin’s side. Bermudez et al published the last series on placental studies in TTTS.11 They included placentas from 26 non-complicated monochorionic– diamniotic pregnancies and from 105 TTTS cases treated by laser. They correlated the fetoscopy findings with an air injection test in order to analyze the photocoagulated areas.

The placentas were classified according to the presence of anastomoses in four groups:
• absence (A)
• presence of only AV anastomoses (B)
• presence of only superficial anastomoses (C)
• presence of both superficial and deep anastomoses (D). 

The highest rate of TTTS was found in group B (85/85). The overall number of anastomoses in TTTS placentas was higher than in noncomplicated monochorionic pregnancies, (4.7 ± 2.48 vs 1.77 ± 1.27, p = 0.000). This study found that deep AV anastomoses were more frequent in TTTS (4.6 ± 2.22 vs 1.4 ± 0.5, p = 0.02), whereas superficial anastomoses were equally similar in both groups (1.6 ± 0.61 vs 1.7 ± 0.96, p = 0.69). The study does not inform which were the number of anastomoses that had to be considered AV or superficial by fetoscopical means. This study contradicted previous studies and has been critically evaluated.12 Laser treatment may alter postnatal injection studies. A referral bias has been mentioned because control placentas were less than one-fifth in number of TTTS placentas. 

Even though there are technical differences in the method, the main conclusion is that the presence of deep AV in the absence of superficial anastomoses are more frequent in TTTS cases. Ex-vivo placental studies are not uniform in selection criteria, and a greater effort should be made in order to select clinically valuable data (Table 4.1). The technique has yet to be uniformly standardized. The findings of different studies are not completely concordant, since one supports the absence of superficial anastomoses as the main conclusion and another supports the absence of superficial AA anastomoses as the main pathological factor. Bajoria and Denbow disagree on the main concept that AV is a critical condition for the occurrence of TTTS. Despite this difference, all workers support the idea that the lack in AA or both AA and VV may generate a circulation system that favors the hypervolemic state of one twin.