Vasa previa is said to occur when fetal vessels, unsupported by placenta or umbilical cord, traverse the membranes over the cervix, below the presenting part. There are 2 variants; type 1 results from velamentous cord insertion and type 2 from vessels running between lobes of a bilobed or succenturiate lobed placenta.1 Spontaneous or artificial rupture of membranes frequently leads to rapid fetal exsanguination.2 As a consequence, the condition is associated with a high perinatal mortality.2–6 Because vasa previa is relatively rare, with a reported incidence of approximately 1 in 2,500 pregnancies,2 most reports in the literature have consisted of 3 or fewer cases. In 1987, Gianopoulos et al7 first described the ultrasonographic diagnosis of vasa previa. Before this, most reports had focused on the universally dismal outcomes with pregnancies complicated by vasa previa.3,4 Since that initial report, several small case series have demonstrated the ability of ultrasonography and color Doppler to diagnose vasa previa prenatally and have suggested improved outcomes associated with prenatal diagnosis of the condition.1,6,8–10
It is intuitive that prevention of perinatal mortality would depend on prenatal diagnosis of the condition and cesarean delivery before rupture of membranes occurs. However, because vasa previa is relatively uncommon, no single institution would be able to put together a large collection of cases.
The primary objectives of this study were to describe pregnancy outcomes in a large series of cases of vasa previa and to assess the differences in outcomes between cases diagnosed prenatally and those not diagnosed prenatally. It was also our objective to determine predictors of neonatal survival in pregnancies complicated by vasa previa.
MATERIALS AND METHODS
This was a chart review of cases of vasa previa. Cases were ascertained from 2 sources, women registering with the Vasa Previa Foundation web site, or through discharge diagnoses from 5 large hospital obstetric services: Georgetown University Hospital, Washington, DC (n = 14); Robert Wood Johnson University Hospital, New Brunswick, New Jersey (n = 3); Saint Peter’s University Hospital, New Brunswick, New Jersey (n = 18); Assaf Harofeh Medical Center, Zerifin, Israel (n = 12); and St George’s Hospital, London, UK (n = 7). At the San Diego Perinatal Center/Sharp Mary Birch Hospital for Women, San Diego, California, cases were ascertained from the perinatal ultrasound database system (n = 14). Appropriate institutional review board approvals were obtained.
The Vasa Previa Foundation (www.vasaprevia.org), established in 2001, consists of women and families who have been affected by pregnancies with vasa previa. Its mission is to raise awareness about vasa previa, to disseminate information on the condition, and to facilitate research aimed at further understanding of vasa previa with the aim of minimizing the perinatal mortality from the condition. The Vasa Previa Foundation provided an invaluable resource for data on pregnancies complicated by vasa previa. Women who registered with the Vasa Previa Foundation web site were invited to participate in the study if they had a pregnancy complicated by vasa previa. After informed consent, their medical records were obtained from their physicians as well as the hospitals where they delivered.
Data were extracted from the charts, including demographic information, information about the affected pregnancy and previous pregnancies, as well as neonatal outcomes. Only cases in which the clinical diagnosis of vasa previa was verified by pathology reports, photographic documentation of the velamentous vessels, or an entry in the chart by the physician documenting the presence of vasa previa at delivery were included. For cases ascertained from the collaborating hospitals, a similar process of verification of vasa previa was applied.
Data were incomplete for some cases regarding location of the placenta in the second trimester (n = 2) and at delivery (n = 2), presence or absence of a bilobed placenta (n = 3), antepartum hemorrhage (n = 4), steroid administration (n = 1), intrapartum hemorrhage (n = 60), mode of delivery (n = 1), and need for transfusion of the newborn (n = 1). Percentages were therefore calculated as proportions of the nonmissing data. In twin pregnancies, only the perinatal data for the first twin was included in our analysis. In all the twin cases studied, the first twin was the fetus whose placenta was associated with a vasa previa. Also, in all the twin cases, the outcomes of the second twin were similar to that of the first. Twenty-five of the cases in this study have been described in previous publications.1,6,11
Statistical analyses were performed with SPSS (SPSS, Inc, Chicago, IL) and Statview for Windows (SAS Institute, Inc, Cary, NC). Comparisons were made between groups based on prenatal diagnosis status and neonatal survival. Because of the potential for bias between the cases ascertained from the Vasa Previa Foundation and those from the participating hospitals, comparisons were also made between these 2 groups based on prenatal diagnosis and neonatal survival. For intergroup comparisons, one-way analysis of variance, Student t test, the χ2 test, Fisher exact test and Mann–Whitney U test were used as appropriate. Statistical significance was defined as a P value of less than .05. Multivariable logistic regression was performed to determine which variables were predictive of neonatal survival. Variables significant in the univariable analysis at a level of P < .2 were considered for inclusion in the regression model. Variables were retained in the model if they remained significant at P < .05.
One hundred fifty-five women with confirmed vasa previa formed the study population. Eighty-seven cases came from the Vasa Previa Foundation and 68 from the participating hospitals. The cases occurred between 1991 and 2003. There were 7 twin pregnancies. In 61 cases (39%), the diagnosis of vasa previa was made prenatally by ultrasonography/color Doppler. Patient demographics compared between those with and without prenatal diagnosis and between neonatal survivors and nonsurvivors are given in Table 1.
Antepartum and intrapartum obstetric factors based on prenatal diagnosis status and neonatal survival status are presented in Table 2. Fifty (32.3%) of the women had bilobed or succenturiate lobed placentas. Ninety-five (61%) had second-trimester low-lying placentas; in only 31 (20%) of these did the placenta remain low-lying at the time of delivery. Fifty-six patients (36.1%) had third-trimester vaginal bleeding. In only 3 cases of third-trimester bleeding, a test was performed to determine whether the blood was of maternal or fetal origin. Of the 108 pregnancies that did not have an elective cesarean delivery, data regarding intrapartum bleeding were available in 91 cases. Of the 87 who did not have prenatal diagnosis, intrapartum bleeding occurred in 82 (94.3%). The mean gestational age at delivery was 34.9 ± 2.1 weeks for cases diagnosed prenatally compared with 38.1 ± 2.5 weeks in cases not diagnosed prenatally.
Perinatal outcomes are reported in Table 3. The overall perinatal mortality was 36%. In 59 (97%) of the pregnancies with prenatal diagnosis, the infants survived, compared with 41 (44%) of 94 when the diagnosis was not made prenatally. Thus the perinatal mortality was 56% when the diagnosis was not made prenatally. Median 1- and 5-minute Apgar scores were 8 and 9, respectively, in cases in which the prenatal diagnosis had been made. However, in the survivors in cases not diagnosed prenatally, the median 1- and 5- minute Apgar scores were 1 and 4, respectively. Twenty-four (58.5%) of the 41 surviving neonates born to mothers whose vasa previa was not diagnosed prenatally required blood transfusions, compared with 2 of the 59 (3.4%) surviving neonates in pregnancies where vasa previa was diagnosed prenatally (P < .001).
In cases from the Vasa Previa Foundation, perinatal mortality occurred in 40 (45.9%) of 87 cases, compared with 15 (22%) of 68 in cases obtained from the hospitals (P = .002). Twenty-five (28.7%) of 87 cases from the Vasa Previa Foundation were diagnosed prenatally compared with 36 (52.9%) of 68 of the cases from the hospitals (P = .02). Of those diagnosed prenatally, the neonatal survival in cases from the Vasa Previa Foundation was 24 (96%) of 25 compared with 35 (97.2%) of 36 cases from the hospitals (P = .99).
On multivariable logistic regression analysis, the variables considered for the model included prenatal diagnosis, prior cesarean delivery, maternal age, administration of steroids for lung maturation, presence of bilobed placentas, in vitro fertilization, the presence of a second-trimester low-lying placenta, a low-lying placenta at the time of delivery, smoking, and gestational age at delivery. Because the source of the data (cases obtained from the Vasa Previa Foundation versus cases obtained from the participating hospitals) may have been a confounder, we adjusted for the data source in our regression model. After removing variables that were not significant (defined as P < .05), the only variables that were significant predictors of survival were prenatal diagnosis (odds ratio 102.9; 95% confidence interval 16.2, 638.3; P < .001) and gestational age at delivery (odds ratio 0.77; 95% confidence interval 0.64, 0.93; P = .01).
Our findings indicate that the best outcomes with vasa previa are achieved when the condition is diagnosed prenatally and the fetus is delivered by cesarean, ideally before the membranes rupture. The most striking finding was that, when the diagnosis was made prenatally, more than 96% of infants survived, whereas more than half of all fetuses/infants died when there was no prenatal diagnosis. Among survivors when the diagnosis had not been made prenatally, 1- and 5-minute Apgar scores were very low (median 1 and 4, respectively). In addition, more than half of surviving neonates required blood transfusions when the diagnosis was not made prenatally.
Previous studies have demonstrated that prenatal diagnosis of vasa previa is feasible. Nomiyama and colleagues,12 using ultrasonography, attempted to routinely locate the insertion of the umbilical cord into the placenta in 667 women. They were able to do this in all cases except 1 and diagnosed 3 cases of vasa previa. These investigators found that it took a mean time of less than 20 seconds to identify the placental cord insertion ultrasonographically and that in 95% of cases it took less than a minute. Lee and colleagues10 and Catanzarite et al,1 using ultrasonography and color Doppler, routinely screened large populations of pregnant women for vasa previa and diagnosed 15 and 10 cases, respectively. More recently, Sepulveda et al13 performed routine sonography to locate the placental cord insertion in 832 women and were successful in 825 cases (99.2%). These authors observed that routinely identifying placental cord insertion added no significant time to the obstetric sonographic examination. It should be emphasized, however, that not all cases of vasa previa necessarily would be recognized by sonography. Such factors as abdominal wall scarring, maternal obesity, or an incompletely filled maternal bladder may prevent visualization of the cord entry into the placenta and/or visualization of fetal vessels over the cervix. Furthermore, vessels that course over the cervix in a transverse rather than an anteroposterior direction may be missed by transabdominal color Doppler ultrasonography.1
Previous studies have suggested that a second-trimester low-lying placenta or placenta previa is associated with vasa previa, regardless of whether the placenta remains low-lying at the time of delivery.5,6,10,14,15 More than 60% of the cases of vasa previa in our study were associated with a second-trimester placenta previa or low-lying placenta, whereas two thirds of these had resolved by the time of delivery. Other investigators have reported that succenturiate and bilobed placentas appear to be associated with the development of vasa previa.1,16 These placental abnormalities were present in 32.9% of our cases, considerably higher than the estimated prevalence in the general population of 4–5%.16,17 Approximately 10% of the women in our study had undergone in vitro fertilization (IVF). It is not clear why IVF appears to be associated with vasa previa.11,14,18 A study of 100 placentas from IVF pregnancies revealed 14 cases of velamentous insertion among them.18 This prevalence was higher than the prevalence of velamentous cord insertion in the general population, even after correcting for the higher prevalence of velamentous insertion in multiple pregnancies.18 Similarly, in a recent study, Schachter and colleagues11 found an incidence of vasa previa at their institution of 1 in 293 IVF deliveries compared with a vasa previa rate of 1 in 6,068 total deliveries. In our study, women with second-trimester low-lying placentas, those with bilobed placentas, and those who experienced third-trimester vaginal bleeding were more likely to have prenatal diagnosis of their vasa previa (Table 2), possibly because the presence of these findings prompted sonographic evaluation of the region overlying the cervix. However, in neither multiple pregnancies nor those resulting from IVF was vasa previa more likely to be diagnosed prenatally.
What should the plan of management be once vasa previa has been diagnosed? Although our study design did not allow us to evaluate the effect of prenatal hospitalization on outcomes, admission to hospital early in the third trimester may be reasonable. Hospitalization allows proximity to facilities for emergent delivery should the membranes rupture. However, we did not evaluate the cost-effectiveness of this approach. An alternative approach is to follow the pregnancies with serial transvaginal cervical length determinations, along with hospitalization should the patient experience contractions or spotting, with a plan for elective delivery at about 35 weeks of gestation. Because of the potential for emergency preterm delivery, consideration should be given to administering steroids to promote fetal lung maturation.
The mean gestational age for delivery in prenatally diagnosed cases was 34.9 ± 2.5 weeks, with 27.9% requiring emergency cesarean delivery because of bleeding, labor, or rupture of membranes, although in the majority of cases elective delivery had been planned for later gestational ages. Based on our findings, we recommend that women with prenatally diagnosed vasa previa be offered elective delivery by cesarean at about 35 weeks of gestation or earlier if fetal lung maturity is documented. This is earlier than the 39 weeks that is generally recommended for elective cesarean delivery. The mean age of delivery for the cases not diagnosed prenatally in our series was 38.2 ± 2.1 weeks. In these cases, perinatal mortality was 56%. Thus, delivery at later gestational ages may negate the benefit of prenatal diagnosis. It is our opinion that the mild risks associated with prematurity at 35 weeks of gestation are outweighed by the risk of a dismal outcome should the membranes rupture, especially because approximately 8% of women at term will have ruptured membranes before the onset of labor.19 Our study demonstrates membrane rupture, even in a hospital, often results in infants with low Apgar scores and who require transfusions, suggesting significant morbidity.
This study is limited by selection bias. Women from the Vasa Previa Foundation had a higher perinatal mortality, which may be related to lower likelihood of having their vasa previa diagnosed prenatally. Because there was an adverse perinatal outcome, these women may have been more likely to seek information on the Internet. Nevertheless, when the prenatal diagnosis was made, there was no statistically significant difference in perinatal survival based on whether cases came from the Vasa Previa Foundation or the hospitals, suggesting that the source of case ascertainment does not appreciably diminish the positive impact of prenatal diagnosis on survival.
This study demonstrates that the mortality from vasa previa can be minimized; a good outcome depends almost entirely on prenatal diagnosis followed by cesarean delivery before rupture of membranes. In very few conditions does prenatal diagnosis lead to such a dramatic improvement in outcome. It is our opinion therefore that standard obstetrical ultrasound protocols should be modified to include screening for vasa previa. An attempt to identify placental umbilical cord insertion should be a routine part of the second-trimester obstetric ultrasound examination. In women at increased risk (those with second-trimester low-lying placentas, pregnancies resulting from IVF, and those with accessory placental lobes), we would recommend consideration of routine transvaginal color Doppler sonography of the region over the cervix if vasa previa cannot be excluded by transabdominal sonography. Based on the findings of this study, it is our opinion that sonographic prenatal diagnosis and careful perinatal management have the potential to prevent the overwhelming majority of fatal outcomes associated with vasa previa.
1. Catanzarite V, Maida C, Thomas W, Mendoza A, Stanco L, Piacquadio KM. Prenatal sonographic diagnosis of vasa previa: ultrasound findings and obstetric outcome in ten cases. Ultrasound Obstet Gynecol 2001;17:109–15.
2. Oyelese KO, Turner M, Lees CC, Campbell S. Vasa previa: an avoidable obstetric tragedy [review]. Obstet Gynecol Surv 1999;54:138–45.
3. Pent D. Vasa previa. Am J Obstet Gynecol 1979;134:151–5.
4. Kouyoumdjian A. Velamentous insertion of the umbilical cord. Obstet Gynecol 1980;56:737–42.
5. Fung TY, Lau TK. Poor perinatal outcome associated with vasa previa: is it preventable? A report of three cases and review of the literature [review]. Ultrasound Obstet Gynecol 1998;12:430–3.
6. Oyelese KO, Schwärzler P, Coates S, Sanusi FA, Hamid R, Campbell S. A strategy for reducing the mortality rate from vasa previa using transvaginal sonography and color Doppler. Ultrasound Obstet Gynecol 1998;12:434–8.
7. Gianopoulos J, Carver T, Tomich PG, Karlman R, Gadwood K. Diagnosis of vasa previa with ultrasonography. Obstet Gynecol 1987;69:488–91.
8. Daly-Jones E, Hollingsworth J, Sepulveda W. Vasa praevia: second trimester diagnosis using colour flow imaging. Br J Obstet Gynaecol 1996;103:284–6.
9. Hata K, Hata T, Fujiwaki R, Ariyuki Y, Manabe A, Kitao M. An accurate antenatal diagnosis of vasa previa with transvaginal color Doppler ultrasonography. Am J Obstet Gynecol 1994;171:265–7.
10. Lee W, Lee VL, Kirk JS, Sloan CT, Smith RS, Comstock CH. Vasa previa: prenatal diagnosis, natural evolution, and clinical outcome. Obstet Gynecol 2000;95:572–6.
11. 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.
12. Nomiyama M, Toyota Y, Kawano H. Antenatal diagnosis of velamentous umbilical cord insertion and vasa previa with color Doppler imaging. Ultrasound Obstet Gynecol 1998;12:426–9.
13. 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.
14. Oyelese Y, Spong C, Fernandez MA, McLaren RA. Second trimester low-lying placenta and in-vitro fertilization? Exclude vasa previa. J Matern Fetal Med. 2000;9:370–2.
15. Francois K, Mayer S, Harris C, Perlow JH. Association of vasa previa at delivery with a history of second-trimester placenta previa. J Reprod Med 2003;48:771–4.
16. Bernischke K, Kaufmann P. Placental shape aberrations. In: Benirschke K, Kaufmann P, editors. Pathology of the human placenta. 4th ed. New York (NY): Springer; 2000. p. 399–418.
17. Fujikura T, Benson RC, Driscoll SG. The bipartite placenta and its clinical features. Am J Obstet Gynecol 1970;107:1013–7.
18. Englert Y, Imbert MC, Van Rosendael E, Belaisch J, Segal L, Feichtinger W, et al. Morphological anomalies in the placentae of IVF pregnancies: preliminary report of a multicentric study. Hum Reprod 1987;2:155–7.
19. Hannah ME, Ohlsson A, Farine D, Hewson SA, Hodnett ED, Myhr TL, et al. Induction of labor compared with expectant management for prelabor rupture of the membranes at term. N Engl J Med 1996;334:1005–10.