Interpretation
Vasa previa is traditionally classified into TypeⅠand Type Ⅱ, whereas more recent papers have described Type Ⅲ. TypeⅠvasa previa is with velamentous cord insertion (VCI), Type II shows interconnecting vessels between two lobes of placenta in a bipartite placenta or connecting vessel with a succenturiate lobe of the placenta[10], and Type Ⅲ refers to boomerang vessels beyond the edges without placental mass protection[11].
The American College of Obstetricians and Gynecologists recommends the usage of color Doppler in patients who are at a tremendous risk for vasa previa[12,13]. TypeⅠ vasa previa will be more easier to be detected by ultrasound, because it just forms at the velamentous part of the cord. Trials have shown Ultrasound performed best for the diagnosis of VCI with a sensitivity, specificity and positive predictive value of 100, 99.9 and 85.7%, respectively, with routine use of color Doppler[14]. TypeⅡvasa previa is more difficult to be diagnosed by ultrasound, because the vasa previa forms between two lobes of placenta with normal umbilical insertion[15].
No large studies have been performed evaluating the effectiveness of MRI in detecting vasa previa[16]. There has been a case study by Kikuchi et al, in which MRI was used as a problem solving measure in identifying the placenta and expediting
management when ultrasonography proved to be inconclusive[17]. But studies have demonstrated MRI may aid in the assessment of placental structures, number of lobes and position, and can play an associated role when diagnosis by ultrasound evaluation is equivocal, especially in case of vasa previa associated with two lobed placentas[18].
In the 12 patients from the study group, MRI demonstrated vasa previa in 11 patients. The rest one was excluded of vasa previa and reached term delivery. In all the 51 patients from the control group, MRI had excluded vasa previa. Pathology examination after delivery had confirmed these MRI diagnoses. The study had shown excellent sensitivity for MRI in diagnosing potential vasa previa when no conclusive ultrasonic result returned, and specificity as well.
The reason we considered that ultrasonic examination had not detected the vasa previa may include the following, but not limited to these reasons:
Firstly, fetal vessels may run at an unfavorable insonation angle to the relatively fixed transducer. The most common sections for pelvic ultrasonography are standard horizontal, sagittal, coronal, and oblique sections originate from the standard ones[19,20]. Thereupon, if the previa vessel long axis was not just in the insonation plane, it would not form a curve image but only a tiny cross section. In this circumstance, the color Doppler could only catch a round red or blue signal in small diameter. Addionally, because of the patients’ low-lying placentas like the 4th, 5th, 7th and 8th paitents, or even with really tiny velamentous parts like 2nd and 3rd patients (Figure 2 & Figure 3), the ultrasound could not distinguish the blood flow of vasa previa from that of the placenta.
Secondly, visualization of vasa previa may be difficult with transvaginal sonography alone. Even if the transvaginal probe actually rotated and remained at the previa vessel long axis direction, the ultrasonic definition of transvaginal gray scale might not be high enough to distinguish the vessel[21]. In this circumstance, the color Doppler could be beneficial, however, the color Doppler scanning should be turned on before the transvaginal probe rotation. Still, due to the low lying placenta making the lower uterine segment not formed well, the previa vessel was not just over the cervical internal os, but posterior to the relatively thick cervical tissue in transvaginal scanning, bringing in further obstacle to identify it, like the 2nd, 4th, 5th, 7th and 8th paitents. Similarly, ultrasound could not clearly exclude vasa previa that with velamentous vessel moving laterally away from cervix as the lower uterine segment extended during the third trimester, like the 12th patient (Figure 12).
Thirdly, in some cases, the vasa previa just carried only a small portion of blood flow, as in TypeⅡand Ⅲ. And recognized from ultrasonologists’ feedback, in other cases, the fetal heads may be so low that compress the previa vessels, making them really difficult be seen. If that happens in TypeⅠvasa previa only with three umbilical vessels in the velamentous part, two umbilical arteries and one umbilical vein, the continuous fetal heart rate monitoring should show fetal distress. But if that is in TypeⅠwith mangrove or even parachute morphology like 4th patient (Figure 4), and merely a tiny branch compressed, or in TypeⅡand Ⅲ, no fetal distress would display. Thus, the blood flow in that previa vessel might be very slow, or not be continuous but intermittent. If there was too low velocity signal that the ultrasonic Doppler sampling gate could not fit well to distinguish it, or even no flow at the time gap during ultrasonic scanning, the vasa previa could still not be identified[22], like 5th and 6th patients (Figure 5 & Figure 6).
Addionally, from this study, it shows that when the vasa previa is close to lower uterine lateral wall, the vessel signal might disguise itself among the uterine blood flow signals. That causes some TypeⅡand Ⅲ cases more difficult to be identified, like 7th to 11th patients.
MRI can be an alternative in identifying vasa previa, if the ultrasound scanning results are inconclusive, owing to some advantages it possesses in this field[23].
Firstly, MRI is independent of the placental location, the maternal obesity, the maternal bladder, the operator, and the directions of velamentous vessels, unlike ultrasound[24]. It may be most useful as an adjunct in difficult cases where ultrasonography is equivocal, the patient is difficult to scan, or if the placenta is implanted in the posterior uterus[25].
Secondly, MRI is tomography, so the image definition for spatial structures and soft tissues will be able to detect absolutely thin vessels[26].
Thirdly, MRI can reconstruct images in three dimensions[27]. A single ultrasonic image only shows one certain plane. While in MRI, the coronal, horizontal and sagittal aspects can be exhibited as 3 simultaneous images within a single display. That assists to determine the vessels positions, within amnion or intramuscular, regardless of different vascular directions. Therefore, any direction of the previa vessels path can be confirmed, not like the ultrasonography that needs the previa vessel’s long axis within the certain scanning plane to see the vessel curve.
Traditional conceptions claimed MRI with several disadvantages, include cost, time, patient monitoring, and the general lack of expertise in interpreting the studies[28]. However, these disadvantages diminished in recent years.
Firstly, MRI costs higher than ultrasound. As the 3-D and 4-D techniques have been widely applied in ultrasonic equipment and elevated the individual scan cost, the financial gap between MRI and ultrasonography is less obvious nowadays[29].
Secondly, MRI is relatively unfit to emergent cases comparing with ultrasound. In real clinical work, if there’s suspected vasa previa with emergent issues such as vaginal bleeding, the most imminent management is cesarean section, not either ultrasound or MRI scanning[30].
Thirdly, MRI is not so available as ultrasound in underdeveloped areas as our country. But accurate ultrasonography diagnosis also needs well experienced and skilled ultrasonologists, which would not be so available in remote hospitals. As more MRI devices widely deployed in recent years, MR scanning might be possible there. Specialists at tertiary centers can analyse the images uploaded from hundreds of miles away. Advanced than ultrasound, MRI digital evidence is fully kept, so any colleagues can review and discuss the previous scanning results at any time when preferred and reach more accurate diagnoses[31].