Since the congenital portosystemic shunt was first discovered by Abernethy[7], this disease has been the subject of an increasing number of studies. But these postnatal classifications refer only to neonates or adults, in-utero classification for fetal congenital portosystemic shunt is still unclear. In view of the unique structure of the UV-PV-DV complex in fetal development, Achiron et al in 2016 proposed the fetal UPSVS classifcation and presented its significance in the prenatal analysis, perinatal management and prognosis prediction[1]. While only a few studies have been conducted on fetal UPSVS, and most are case reports or retrospective studies with small sample sizes currently, the characteristics, pregnancy outcomes and follow-up results of UPSVS has not been clearly studied in fetal period. For this reason, we applied the new UPSVS classification to analyze 40 cases of fetal UPSVS in our center, and discovered that: 1) the majority of cases had Type II and Type IIIa shuns; 2) cases with Type IIIb (2/2) had the worst outcome-pregnancy of termination (TOP), followed by Type I (2/4); 3) Type I (3/4) and Type IIIa (9/16) UPSVS was associated with cardiac enlargement, Type II was characterized by structural malformations, Type IIIa was associated with the highest incidence (8/16) of FGR; 4) spontaneous shunt closure occurred in all survivors of Type IIIa.
In the present study, we discovered that UPSVS Type II and Type IIIa was detected in the majority of cases following the new classification. This postnatal result is in line with this of previously published finding, suggesting that type II and type IIIa UPSVS prevalence could be higher. Achiron and Kivilevitch retrospectively analyzed 44 cases of prenatally diagnosed UPSVS and found that 19 neonates (43.18%) had DVSS and 12 neonates (27.27%) IHPSS[1].
UPSVS has always been associated with multiple congenital abnormalities. According to our data, the most common abnormalities related to UPSVS took place in the cardiovascular system, which is consistent with previous report[8]. The Type II (DVSS) group was associated with structural malformations, including the PLSVC, ARSA, TOF and so on. Type IIIa shunt (IHPSS) was characterized by cardiac enlargement. But the two groups always had the highest ratio of a marvelous outcome due to a normal PV system. The reason why cases with UPSVS had abnormal cardiovascular development may be that the abnormal persistence of an embryonic vitelline vein may lead to hemodynamic abnormalities, which be secondary to congenital heart disease[9]. Another probable mechanism is that increased cardiac preload caused by abnormal shunts, thus lead to enlargement of the heart, which suggests that detailed scanning for cardiac structure and function should be required in UPSVS cases[10]. FGR is the secondary abnormality and of note, all cases associated with FGR were IHPSS. The finding is in agreement with previous studys pointing out that FGR has the highest incidence (47.4%) of intrauterine concomitant abnormality in Type IIIa of shunts[10]. One of the possible reasons lies in abnormal shunting of the PV and the HV shunts cause liver blood perfusion to reduction, subsequently leading to a decrease in foetal insulin growth factors, therefore, peripheral tissue growth impairs[11, 12]. These results suggested that appropriate attention should be given to screening fetal growth and development in the cases with type IIIa.
Our results showed that TOP in 2 type I cases and 2 type IIIb cases was determined by the completely absence of intrahepatic PV system; TOP in 4 type II cases and 6 type IIIa cases was requested by their family members; additionally, 1 case of Type IIIa resulted from premature at 37th week gestation. It has been observed that USS has a high incidence of mortality in all UPSVS types, in line with our findings[13]. The lowest live birth rates and the poorest prognosis for EHPSS was also reported in a previous study[4]. We noted poor development of the intrahepatic PV system in these USS and EHPSS cases, which may be the determining factor for dismal prognosis. Meanwhile, we should understand that there are still other situations, including TOP in some cases was not decided by the shunt but rather requested by the couples.
UPSVS is also associated with abnormal genetic background, such as trisomy 21, trisomy 13 and so on[14]. In our study, one case of Type I was diagnosed with q22.11 deletion in trisomy 21 and one case of Type IIIa q21.1 duplication in trisomy X. In contrast, Haifang Wu et al revealed abnormal karyotypes in all three types of shunts[13]. Dong X et al thought Type I and Type II UPSVS cases have a high rate of abnormal results in genetic tests[15]. Achiron and Kivilevitch discovered that chromosomal abnormalities only could be found in Type 2 shunts, but no in Type I and Type III shunts[1]. Since these results are based on small sample studies and the methods of genetic analysis are also different, whether these variations or inversions in genetic results were associated with the type of UPSVS or were noticed occasionally requires further research. On balance, genetic abnormalities might occur in any of the three types of UPSVS and prenatal genetic analysis sometimes are needed.
We observed spontaneous complete closure during follow-up observation in all Type IIIa survivors with the best outcome, which is similar to the conclusion of previous studys. Han et al found spontaneous shunt closure occurred in most survivors (5/6) during a period of 1–12 months[16]. Zhu et al retrospectively studied 10 fetuses with IHPSS and showed that half of the survivors had complete and partial spontaneous shunt closure at approximately 16 months after birth[17]. Although sometimes medical interventions may be not necessary for this type, surgical treatment become important choice for neonates whose shunt has not spontaneous closure after 2 years or who are symptomatic[18].
Limitations
This study had some limitations. Firstly, neither prenatal ultrasound findings had been confirmed nor fetal specimens collected for later fetatology tests after abortion, as this was a retrospective study. Secondly, the follow-up time is short, and there is no long-term follow-up.