Vasa previa has been defined as the presence of a fetal vessel or vessels unsupported by either placenta or umbilical cord running in the membranes below the presenting fetal part, either over the cervix1,2 or in the lower (unquantified) uterus.3–7 In reviewing the literature in 1999, Oyelese5 found an overall incidence of 1 in 2,500 births in four major English language series, totaling 24 cases of vasa previa with a range of 1 in 1,275 to 1 in 8,333. More recent studies support this range.2,3,6,8 Risk factors include velamentous cord insertions, the presence of accessory placental lobes, a low-lying placenta or placenta previa earlier in the pregnancy, twins, and in vitro fertilization.3,8,9
Vasa previa has been associated with perinatal mortality rates of 0–36% in recent series.1–3,6 Prenatal diagnosis has been associated with improved neonatal outcome.1,6,10 High prenatal detection rates (83–100%) have been reported in smaller series and in those with a set protocol for evaluation.2,3,9 Once diagnosed, timing of delivery is not uniformly agreed on.1,6,10,11 The current literature regarding vasa previa predominately consists of relatively small series and case reports, limiting recommendations for management. In a review of 48 cases published from 1980 to 1997 (plus three of their own), Fung10 recommended elective delivery before the onset of labor or rupture of membranes. In more recent series, both Smorgick (19 cases) and Oyelese (155 cases) reported that prenatal diagnosis (mean gestational age at delivery of 35 weeks) was associated with better neonatal outcomes than postnatal diagnosis (mean of 37 weeks to 38 weeks of gestation at delivery).1,6 In one pooled analysis two factors, diagnosis made prenatally and gestational age at delivery, were predictive of survival.1 On the basis of these studies, 35 or 36 weeks of gestation has been deemed a reasonable gestational age for elective delivery.12 However, in a physician survey conducted in England and Wales, only 1% of the respondents stated that they would perform elective delivery at 35 weeks of gestation or earlier, whereas 75% stated they would deliver at 37 or 38 weeks of gestation.11
Yet to be addressed in the vasa previa literature are the optimal management for twins, the clinical significance of a fetal vessel, ie, near but not directly over the cervix, and gestational age-specific risks. To evaluate clinical issues involved with vasa previa, we performed a two-decade review of case group participants delivered at our hospital.9
MATERIALS AND METHODS
The study is a retrospective review of cases of vasa previa in women who delivered at William Beaumont Hospital, Royal Oak, Michigan from January 1, 1990, to June 30, 2010, and it was approved by the Beaumont Hospital Human Investigation Committee. Cases were identified from two sources: the electronic files for all ultrasound scans performed in the Division of Fetal Imaging and hospital discharge records.
Our ultrasound examinations are performed by board-certified sonologists and were reviewed by physicians certified by either Maternal-Fetal Medicine or Radiology boards. All ultrasound examinations performed at 15 weeks of gestation and later in which a vasa previa, defined as a fetal vessel within the membranes of the lower uterus, was reported were used for this study. Our routine ultrasound protocol included a two-dimensional evaluation of the lower uterus and cervix. Patients were instructed to arrive with a full bladder to assist in the visualization of the lower uterus and cervix. Identification of the umbilical cord insertion site was performed on a routine basis for twins. Transvaginal ultrasonography and color Doppler evaluation were not part of our routine protocol but were used when needed to assist in the pregnancy evaluation. In case group participants in whom the diagnosis of vasa previa was not made on earlier ultrasonogram, and in case group participants in whom the vasa previa resolved on subsequent ultrasonogram, video recordings of the ultrasonogram (video cassettes until 2005 and digital clips from 2005 onward) were retrospectively reviewed.
A vasa previa was diagnosed when a fetal vessel or vessels were noted in the lower uterus. The location of the membranous fetal vessel was classified as overlying the cervix or low-lying when present in the lower uterus (a yet to be quantified term) but not directly over the cervix. When low-lying, the distance from the membranous vessel to the internal os was measured. Women with cases of vasa previa that were diagnosed in our unit but who delivered elsewhere were excluded. Cases in which the internal cervical os was covered by placenta, regardless of gestational age, were classified as placenta previa. A low-lying placenta was indentified when the lower placental edge was within 2 cm but not covering the internal cervical os.
Hospital records were reviewed for prenatal, delivery, and neonatal information. Prenatal data recorded included the presence and gestational age of antenatal bleeding episodes. Delivery information included gestational age and clinical circumstances at the time of delivery and description of the placenta and umbilical cord insertion. Deliveries were classified as emergent, urgent, or elective. Emergent deliveries were defined as those performed for bleeding associated with nonreassuring fetal heart rate patterns. Elective deliveries were defined as those that were scheduled for the diagnosis of vasa previa without ongoing acute maternal symptoms or evidence for fetal compromise. Urgent deliveries were defined as those with an acute ongoing complication (bleeding, labor, rupture of membranes) in which both the maternal and fetal statuses were otherwise clinically stable. Neonatal information recorded included umbilical pH values, Apgar scores, and need for transfusion.
For statistical analysis, continuous variables (gestational ages at initial bleed and delivery) were summarized using mean±standard deviation, minimum, median, and maximum, whereas categorical variables (pregnancy type, vessel location, and indication for delivery) were summarized using frequencies and percentages. Data were first assessed using numerical and graphical techniques to determine whether they met the distributional assumptions of the statistical tests being used to analyze them. Based on this preliminary assessment, categorical variables were compared using the two-tailed Fisher exact test, whereas continuous variables were compared using the Wilcoxon two-sample test using the t approximation. Statistical analysis was performed using the SAS System for Windows 9.2. P<.05 was considered statistically significant.
Over the study period, 81 cases of vasa previa were diagnosed through either prenatal ultrasound examinations or placental evaluations after delivery. Five of the women with cases diagnosed by ultrasonogram had subsequent ultrasound examinations and delivery information that did not support the diagnosis of a vasa previa. In two of the women, ultrasound examinations were not available for our review. Given that there were 182,554 pregnancies that had at least one ultrasound examination in our unit over the course of the study period, our incidence of vasa previa was 1 in every 2,467 pregnancies.
Excluding 16 case group participants who were referred to us only for ultrasound evaluation and had their prenatal care and delivery elsewhere, and excluding the five false-positive diagnoses, there were 60 case group participants (seven twin and 53 singleton pregnancies) in our study group.
In the 58 case group participants who had an ultrasonogram in our unit, the placenta was either low-lying or overlying the cervix after 14 weeks of gestation in 36 case group women (62.1%). The umbilical cord insertion site was not routinely evaluated on our prenatal ultrasonograms; however, but in 48 case group participants, delivery and placental pathology reports allowed for evaluation of the cord insertion site. A velamentous or marginal cord insertion was present in 43 (89.6%) women. The presence of either an abnormal cord insertion (marginal or velamentous) or an abnormal placental location (low or over the cervix) was present in 55 of the 58 case group women (94.8%).
Overall, our diagnosis of vasa previa was made before delivery in 56 of the 58 case group women (96.6%). In the two patients in whom the diagnosis was not made, both had only single 19-week ultrasonograms. In 42 case group participants, the fetal vessel was located over the cervix; in 14 case group participants, the fetal vessel was low-lying (defined as located near but not directly over the cervix). In all but one of these case group women the low-lying vessel was within 3 cm of the internal os. On our initial ultrasound examination after 14 weeks of gestation, we correctly identified the vasa previa in 26 cases and missed the diagnosis in eight cases. In the remaining 24 case group participants, placenta previa was present, precluding a diagnosis of vasa previa at that time. On the first ultrasound examination after resolution of the placenta previa, we identified the vasa previa in 19 case group participants and missed it in five case group participants.
On our initial ultrasonogram without a placenta covering the cervix, the diagnosis of vasa previa was correctly made in 45 case group participants (77.6%). To provide insight into our missed diagnoses, we retrospectively reviewed the digital clips and video cassette tapes from the 13 case group participants in whom the vasa previa was not detected on our initial ultrasonogram without a covering placenta. Abdominal ultrasonograms were performed for all 13 patients. Linear or curvilinear structures along the membranes of the lower uterus, suggestive for vasa previa but not noted on the initial reading of the ultrasonogram, were seen in four case group women; in six case group women, no suspicious areas for vasa previa were seen despite adequate visualization of the lower uterus; in three cases, visualization of the lower uterus was thought to be inadequate to evaluate for vasa previa. Seven of these case group participants also had a transvaginal ultrasonogram performed, typically for cervical length or evaluation of placental location. Six of the seven women had findings that were suspicious for vasa previa but were not seen on the initial reading of the examination, whereas in the seventh woman no evidence for vasa previa was seen. In all but two case group participants, the diagnosis of a vasa previa was made later in pregnancy on subsequent examinations.
In five cases of diagnosed vasa previa, subsequent ultrasonograms failed to detect any vessels near the cervix, and a normal placenta and cord insertion were noted on delivery. On retrospective review, in one case the “vasa previa” was likely a loop of umbilical cord in the lower uterus (funic presentation), and in the other four cases the vessels seen were likely dilated maternal sinuses or vessels on the internal surface of the cervix and lower uterus.
There were three perinatal deaths, two neonatal and one antenatal, for a mortality rate of 4.5% (3/67 fetuses, including twins). Emergent deliveries occurred in seven cases (11.7%) and were associated with significant neonatal morbidity and mortality (Table 1). Six of the newborns had hypovolemic shock at birth requiring intubation and neonatal intensive care unit care, whereas the seventh was stillborn and failed to respond to resuscitative measures. All neonatal deaths occurred in this group. Neonatal outcomes from the nonemergent deliveries were favorable. No case involved an umbilical artery pH less than 7.2, and only two neonates in this group required transfusions, both in the third week of life. Antepartum bleeding was relatively common, occurring in 25 (42%) case group participants. The mean gestational age at the time of the initial bleed was 30.3 weeks of gestation (±4.2 weeks). Six of these case group participants were delivered with their initial bleeding episode, two emergently. The other 19 delivered, on average, 21.1 days later (±24.4 days; range 2–73 days). Two of these were emergent deliveries. The percentage of patients remaining undelivered at each gestational age for the entire study group is shown in Figure 1.
In four case group participants, all with singleton pregnancies, the diagnosis of vasa previa was not made before delivery. Three of these case group participants had an emergent delivery while in labor at term (39 weeks, 41 weeks, and 41 weeks of gestation). The fourth woman with an undiagnosed case was delivered with recurrent antepartum bleeding and normal fetal heart rate tracings, with a presumed diagnosis of recurrent abruption. The 75% emergent delivery rate with undiagnosed cases was significantly increased compared with the 7% seen in cases diagnosed prenatally (P=.002).
Overall indications for delivery (n, %) in the seven twin pairs were elective (three [42.9%]), emergent (two [28.6%]), and urgent (two [28.6%]). Indications for delivery in singletons were elective (28 [52.8%]), emergent (five [9.4%]), and urgent (bleeding with normal fetal heart race tracings in nine [17.0%]; labor or premature rupture of membranes in 11 [20.8%]).
To evaluate the clinical significance of pregnancy type (singleton compared with twin) and vessel location (low-lying compared with over the cervix), the emergent delivery rates were compared (Table 2). Twin pregnancies had a significantly earlier median age at delivery of 32 weeks of gestation as compared with 35 weeks of gestation in singletons (P=.01). All twins were delivered by the completion of the 35 weeks of gestation, with emergent deliveries at 27 and 31 weeks of gestation. Gestational age–specific risks for delivery are illustrated in Figure 2. The chances of delivery over the course of the next week, both total and emergent, are shown for all case group women still pregnant at the start of each gestational age week. Case group women with a low-lying vasa previa had emergent delivery rates that were higher than those seen with the fetal vessel over the cervix (Table 2), but this difference also was not significant. In all emergent deliveries, the membranous fetal vessel was within 2 cm of the internal os (obviously undiagnosed cases are not included). In the lone vaginal delivery in the group (uncomplicated), a vasa previa noted at 20 weeks of gestation was not seen on a 27-week follow-up ultrasonogram. Our retrospective review of the 27-week ultrasound examination showed a membranous cord insertion 2.5 cm from the internal os that was initially overlooked. The membranous cord insertion was confirmed on placental pathology.
This study evaluated the diagnosis of vasa previa, along with the significance of twins and vessel location, and calculated the weekly gestational age–specific risks for emergent delivery. Risk factors were useful in screening for vasa previa cases, with the combination of either an abnormal placental location (either low-lying or over the cervix) or a velamentous or marginal cord insertion (noted here on delivery) present in 55 of the 58 case group women who had ultrasonograms in our unit. Other studies also have shown high detection rates with specific protocols that include umbilical cord insertion identification and lower uterus and cervix evaluation.2,3
Two observations came from our retrospective evaluation of the case group individuals in whom we failed to diagnosis the vasa previa. First, transabdominal ultrasonograms were less likely to detect evidence of the vasa previa. Second, transvaginal examinations, performed typically for either cervical length or placental location, often had suspicious areas on retrospective review that were initially overlooked, suggesting those performing the ultrasonograms may have been overly focused on the cervix and placental edge, because the ultrasonograms were ordered to evaluate these structures. To optimize the detection of vasa previa, transvaginal evaluation of the lower uterus is recommended, for at least the at-risk case group individuals.4 Whenever a transvaginal ultrasonogram is performed, the lower uterus should be evaluated, looking specifically for a fetal vessel. Doppler interrogation of suspected membranous vessels in the lower uterus can be useful in limiting incorrect diagnoses.2 Additionally, the possibility of a funic presentation should be excluded.
Once vasa previa is diagnosed, the decision regarding delivery timing must balance the risks of prematurity with the potential for fetal bleeding resulting from a delayed delivery. Antepartum bleeding was a common finding (42% of the case group participants), but delivery did not occur at the time of the initial bleed in most instances. With careful monitoring of the fetal status, increased gestational age can be gained. Unfortunately, because these case group women were managed without a unifying protocol, we are unable to provide specific recommendations for delivery timing in cases of bleeding and reassuring fetal testing.
Improved neonatal outcome has been associated with prenatal diagnosis and gestational age at delivery, and delivery at 35 weeks of gestation has been recommended.1,6,12 Low-lying vasa previas had similar emergent delivery rates as cases of the fetal vessel over the cervix. Diagnostic criteria for a low-lying vasa previa cannot be definitely set at this time because of limited numbers, although it should be noted that all of our emergent deliveries had the fetal vessel within 2 cm of the cervix.
In singletons, our overall risk of emergent delivery was 9.4%, attributable in part to the three undiagnosed cases in women who were in labor at term (39 weeks, 41 weeks, and 41 weeks of gestation). Excluding these, our preterm emergent delivery rate for singletons was approximately 4%. In singletons, clearly the risk of continuing pregnancy once 39 weeks of gestation has been reached is offset by the increased risk of emergent delivery (Fig. 2). Before this, our data did not identify an obvious choice for timing of delivery.
There are several limitations to our study. The small number of twin cases and cases of low-lying vasa previas is an obvious one. We cannot exclude the possibility that the low-lying vasa previa cases were managed differently than cases of the vasa previa located over the cervix, increasing the chance of an adverse outcome for the former. Additionally, because placental examination is not universal in our institution, we almost certainly have underascertained cases of vasa previa and, to the extent that we did, our findings may not reflect the true clinical implications of vasa previa. Questions regarding other potentially important clinical factors, such as the type of fetal vessel involved (membranous cord insertion at or near the cervix compared with a distal branch of a fetal vein near the cervix), the role of routine hospitalization,9 and cervical length measurements, cannot be answered by our data.
1. Oyelese Y, Catanzarite V, Prefumo F, Lashley S, Schachter M, Tovbin Y, et al.. Vasa previa: the Impact of prenatal diagnosis on outcome. Obstet Gynecol 2004;103:937–42.
2. Catanzarite V, Maida C, Thomas W, Mendoza A, Stanco L, Piacquadio KM. Prenatal sonographic diagnosis of casa previa: ultrasound findings and obstetric outcome in ten cases. Ultrasound Obstet Gynecol 2001;18:109–15.
3. Baulies S, Maiz N, Muñoz A, Torrents M, Echevarría M, Serra B. Prenatal ultrasound diagnosis of vasa praevia and analysis of risk factors. Prenat Diagn 2007;27:595–9.
4. Cipriano LE, Barth WH Jr, Zaric GS. The cost-effectiveness of targeted or universal screening for vasa praevia at 18-20 weeks of gestation in Ontario. BJOG 2010;117:1108–18.
5. Oyelese KO, Turner M, Lees C, Campbell S. Vasa previa: an avoidable obstetric tragedy. Obstet Gynecol Surv 1999;54:138–45.
6. Smorgick N, Tovbin Y, Ushakov F, Vaknin Z, Barzilay B, Herman A, et al.. Is neonatal risk from vasa previa preventable? The 20-year experience from a single medical center. J Clin Ultrasound 2010;38:118–22.
7. Sepulveda W, Rojas I, Robert JA, Schnapp C, Alcalde JL. Prenatal detection of velamentous insertion of the umbilical cord: a prospective color Doppler ultrasound study. Ultrasound Obstet Gynecol 2003;21:564–9.
8. Schachter M, Tovbin Y, Arieli S, Friedler S, Ron-El R, Sherman D. In vitro fertilization is a risk factor for vasa previa. Fertil Steril 2002;78:642–3.
9. Lee W, Lee VL, Kirk JS, Sloan CT, Smith RS, Comstock CH. Vasa previa: prenatal diagnosis, natural evolution, and clinical outcome. Obstet Gynecol 2004;95:572–6.
10. Fung TY, Lau TK. Poor perinatal outcome associated with vasa previa: is it preventable? A report of three cases and review of the literature. Ultrasound Obstet Gynecaol 1998;12:430–3.
11. Ioannou C, Wayne C. Diagnosis and management of vasa previa: a questionnaire survey. Ultrasound Obstet Gynecol 2010;35:205–9.
© 2013 by The American College of Obstetricians and Gynecologists.
12. Oyelese Y, Smulian JC. Placenta previa, placenta accreta, and vasa previa. Obstet Gynecol 2006;107:927–41.