Background
Vasa previa is a rare condition that is associated with a high rate of fetal or neonatal death when not diagnosed antenatally. The majority of available studies are either small, do not include antepartum data, limited to single institutions, or are biased by inclusion of patients from registries and online vasa previa support groups.
Objective
The purpose of this study was to investigate the diagnostic and management strategies for this potentially catastrophic entity and to describe further maternal and placental risk factors that may aid in the establishment of a screening protocol for vasa previa.
Study Design
This was a retrospective multicenter descriptive study that included all pregnancies that were complicated by vasa previa that delivered between January 1, 2000, and December 31, 2012. Nine maternal fetal medicine practices and the hospitals in which they practice participated in data collection of diagnosis, treatment, and maternal-neonatal outcomes.
Results
Sixty-eight pregnancies were identified that included the diagnosis of vasa previa or “possible vasa previa” either in the ultrasound record or in the hospital record at the time of delivery. Four cases (5.8%) appeared to resolve on repeat ultrasound examination. Fifteen of the 64 cases that were suspected of having vasa previa could not be verified or were not documented at delivery. Of the remaining 49 cases, where vasa previa was documented, 47 cases (96%) were diagnosed by ultrasound scanning antenatally. Known risk factors for vasa previa were present in 41 of 47 cases (87%). Of the 49 cases, 41 were delivered by planned cesarean delivery at a mean gestational age of 34.7 weeks, and 8 cases required emergent cesarean delivery at a mean gestational age of 34.6 weeks (range, 32.4-36.0 weeks gestation). Seven of these emergent cesarean deliveries had been diagnosed previously; 1 case had not. All of the emergent cesarean deliveries were for vaginal bleeding; 1 case was also for a concerning fetal heart rate, but only 1 of the known cases had a documented ruptured fetal vessel. None of these cases were found to have cervical shortening before the onset of bleeding. One of the undiagnosed cases resulted in a ruptured fetal vessel and a baby with no heart beat at birth who survived but had periventricular leukomalacia at 1 month of age with mild white-matter atrophy. Of the remaining neonates in this group, there were no deaths and no major complications beyond mild respiratory distress syndrome in 9 cases. There were no other major neonatal complications, which included no cases of periventricular leukomalacia, neonatal sepsis, necrotizing enterocolitis, or any grade of intraventricular hemorrhage in the confirmed cases of vasa previa.
Conclusion
This study confirms most current recommendations that include risk-based ultrasound screening, early hospitalization at 30-34 weeks gestation, antenatal corticosteroids at 30-32 weeks gestation, and elective delivery at 33-34 weeks gestation. Thus, with these recommendations for current identification and management of vasa previa in this series of geographically diverse mostly private practice maternal fetal medicine practices, we have confirmed recent reports that show a dramatic improvement in neonatal survival and complications compared with earlier reports.
Few antenatal diagnoses carry the burden and fear of inherent and unanticipated lethality as vasa previa. The unprotected fetal vessels traversing freely within the amniotic membrane over the internal cervical os are vulnerable to both spontaneous and iatrogenic rupture as the cervix shortens and labor ensues. On the rupture of these unprotected fetal vessels, which most often occurs at the time of membrane rupture, the fetus can exsanguinate within minutes, often sooner than cesarean delivery can be accomplished. Fortunately this diagnosis is rare, with a reported incidence of 1:500 to 1:6000 pregnancies.
The antenatal clinical diagnosis of vasa previa can be missed without careful attention to routine ultrasound guidelines and requires ultrasound scanning for the identification of the cord insertion, careful assessment of the lower uterine segment, and the use of color Doppler ultrasound scans. Organizations that often determine standard of practice, which include the American Congress of Obstetricians and Gynecologists and the American Institute of Ultrasound in Medicine, do not include a recommendation for routine screening for vasa previa as part of the complete ultrasound examination. Despite this, the antenatal recognition of the rare entity is crucial, given the nearly universally fatal consequences of fetal vessel rupture with vasa previa. Reports in the literature are quite promising when the diagnosis of vasa previa are made antenatally, with survival rates of 97-99%; however, if the diagnosis is missed, the survival rate drops significantly. If an emergency cesarean delivery is required and if the diagnosis of vasa previa is not made in the antenatal period, <50% of neonates survive. Therefore, it is critical that the diagnosis of vasa previa be made before presentation for labor or prelabor rupture of membranes.
It is arguable whether universal screening should be advocated or would be cost-effective. Currently, the most agreed on recommendation is a risk-based approach to ultrasound screening for vasa previa based on the identification of maternal and placental risk factors that have been reported to include placenta previa, low-lying placenta, succenturiate placenta or an accessory placental lobe, advanced maternal age, multiple gestation, and in vitro fertilization. Once the diagnosis is made antenatally, there is no clear consensus on the best approach to management, which includes the need for antepartum admission; the need for and timing of antenatal corticosteroids (ACSs); the value of antepartum fetal heart rate testing; in particular, the timing of elective cesarean delivery; and whether factors like parity, pregnancy history, and/or cervical length measurement should enter into the decision for the appropriate timing of delivery.
Current literature is also limited in that the largest study is derived at least in part from registries and online support groups of women who previously have been afflicted with vasa previa. Databases and case series that have reported primarily on such cases are likely to be biased towards poorer outcomes, given that women with negative outcomes are more likely to seek support groups and membership in online registries. Even data bases are limited in that the clear reporting of this entity is also more likely to be done when a more dramatic or adverse outcome occurs. One other large study of 58 cases from a single center over a 20-year period provides encouraging results regarding the improvement in antenatal ultrasound diagnosis and improvement in neonatal outcome. However, the establishment of the success of the identification and avoidance of the potentially lethal perinatal outcome in this rare, but potentially lethal, entity in routine practice is the ultimate test of whether real progress has been made. Therefore, with the establishment and availability of a large collaborative research group of maternal fetal medicine practices with access to ultrasound and hospital data bases, we sought to identify cases and review the data from this multicenter group by collecting information that might be helpful in elucidating the optimal diagnosis and management approaches for vasa previa.
Material and Methods
The protocol and a waiver of consent were approved by central and/or local institutional review boards for each participating site. Physicians and research nurses from 9 maternal fetal medicine practices and the hospitals in which they practice participated in collecting and reviewing the data. This collaborative research group is composed of maternal fetal medicine practices managed by the Pediatrix/Obstetrix Medical Group and 1 other university practice with an interest in participating in collaborative clinical research studies. All sites had ≥1 dedicated obstetric research nurses and a designated investigator for each site. There were no uniform diagnostic or management approaches because all units function autonomously in clinical decision-making and because none had defined protocols for vasa previa. Screening all patients for vasa previa with endovaginal ultrasound scans was not routine in any of the practices. Databases from ultrasound scans and hospital records of patients who delivered from January 1, 2000, through December 31, 2012, were reviewed. Patients with the diagnosis of vasa previa or with a suspicion of vasa previa on ultrasound scanning that was made antenatally or at the time of delivery via clinical and/or pathologic investigation were reviewed for inclusion. Once a patient was identified as having met the study’s inclusion criteria, a unique study number was assigned. All pertinent ultrasound reports, maternal and neonatal hospital records, operative reports, ultrasound reports, and placental pathology reports were reviewed. No effort was made to further define diagnostic descriptions of findings on the ultrasound reports such as placenta previa or other placental abnormalities, because the details of these findings were often variable or completely omitted. Gestational age was defined on the basis of the ultrasound reports. Deidentified data were then entered on specifically designed and institutional review board–approved case report forms that were coded by the unique identification number. Only those patients with a specific description of vasa previa in the operative or delivery report and/or on pathology report of the placenta met the criteria for the final diagnosis of confirmed vasa previa. The operative or delivery report required a description of visualization of fetal vessels traversing the membranes over or adjacent to the cervical canal; the pathology report required a cord insertion at or beyond the margin of the placenta with the fetal vessels running freely in the membranes. In addition, we reviewed whether there was disruption of a fetal artery or vein or both in either report.
Results
Sixty-eight pregnancies were identified that included the diagnosis of vasa previa or “possible vasa previa” either in the ultrasound record or in the hospital record at the time of delivery. Four of the antenatally diagnosed cases (5.8%) appeared to resolve on repeat ultrasound examination and did not present a problem; they were not identified as vasa previa at the time of delivery. Resolution was documented in these 4 cases at 21, 32 (2 cases), and 33 weeks gestation. Fifteen of the 64 cases that were suspected of having vasa previa at the time of delivery could not be verified by the operative/delivery report or by a pathology report. Of the remaining 49 confirmed cases, for which vasa previa was documented either on the operative report, the pathology report or both, 47 cases (96%) were diagnosed by ultrasound scanning antenatally, and 2 cases (4%) were documented at delivery only. Maternal and placental risk factors and characteristics are presented in Table 1 . Most of these cases (86%) had an identifiable placental and/or maternal risk factor. The confirmed cases delivered at a mean gestational age of 34.7 weeks ( Table 2 ). Eight patients required emergent cesarean delivery at a mean gestational age of 34.6 weeks (range, 32.4–36.0). Seven of these cases had been diagnosed previously, and 1 case had not ( Table 3 ). All of these emergent cesarean deliveries were for vaginal bleeding, and one also for a concerning fetal heart rate. Six of the 8 deliveries had a cervical length measured by endovaginal ultrasound scanning (all but 1 within a week of the emergent delivery; Table 3 ), and no cervical length was significantly shortened before the emergency delivery. One emergent delivery was in the patient whose vasa previa had not been diagnosed antenatally. The mother presented at 36 weeks with labor, membrane rupture, vaginal bleeding, and a category II fetal heart rate pattern. This baby had Apgar scores of 0, 0, and 2 at 1, 5, and 10 minutes, respectively, with a cord arterial pH of 6.55 and an initial hematocrit level of 26.7%. The neonate ultimately survived but had periventricular leukomalacia at 1 month of age with mild white matter atrophy. The second undiagnosed patient was delivered by elective repeat cesarean delivery; a vasa previa was visualized and documented in the operative report. Details of these emergency cesarean deliveries are provided in Table 3 . Of the remaining 41 cases, all underwent planned cesarean delivery at a mean gestational age of 34.7 weeks. A total of 37 of 49 cases (76%) received either 1 or 2 courses of antenatal steroids. Five patients (10%) underwent amniocentesis to determine fetal lung maturity before delivery.
| Characteristics | Ultrasound scan diagnosed and confirmed (n = 47) | Ultrasound scan diagnosed and not confirmed (n = 13) | Not diagnosed on ultrasound scan (n = 2) | Ultrasound scan diagnosed and resolved (n = 4) |
|---|---|---|---|---|
| Placental | ||||
| Color Doppler imaging used, n (%) | 34 (72.3) | 13 (100) | 1 (50.0) | 1 (25.0) |
| Transvaginal ultrasound scanning used, n (%) | 39 (83.0) | 12 (92.3) | 1 (50.0) | 4 (100) |
| Initial cervical length, cm a | 3.9 ± 0.67 | 4.1 ± 1.08 | 3.8 ± NA | 4.0 ± 0.46 |
| Shortest cervical length, cm a | 3.3 ± 0.81 | 3.7 ± 1.04 | 2.4 ± NA | 2.9 ± 1.03 |
| Low lying placenta, n/N (%) | 24/45 (53.3) | 7/11 (63.6) | 0 | 1/4 (25.0) |
| Marginal cord insertion, n/N (%) | 12/46 (26.1) | 0/11 (0) | 0 | 2/4 (50.0) |
| Velamentous umbilical cord insertion, n/N (%) | 20/46 (43.5) | 2/11 (18.2) | 0 | 2/4 (50.0) |
| Bilobed or sucenturiate lobed placenta, n/N (%) | 17/46 (37.0) | 4/11 (36.4) | 0 | 1/4 (50.0) |
| Mean gestational age at detection of vasa previa b | 27.0 (16-40) | 27.8 (16-32) | 19.8 (18-21) | |
| Maternal | ||||
| Age ≥35 y, n (%) | 20 (42.6) | 4 (30.8) | 0 | 0 |
| In vitro fertilization, n (%) | 8/44 (18.2) | 1/11 (9.1) | 0 | 0 |
| Multiple gestation, n (%) | 5 (10.6) | 0 | 0 | 0 |
| No risk factors for vasa previa, n (%) c | 6 (12.8) | 1 (7.7) | 0 | 3 (75.0) |
a Data are given as mean ± standard deviation
b Data are given as median (range)
c If a risk factor was missing, it was counted as “No” for the “No Risk” factor for vasa previa.
| Variable | Ultrasound scan diagnosed and confirmed (n = 47) | Ultrasound scan diagnosed and not confirmed (n = 13) | Ultrasound scan diagnosed and resolved (n = 4) | Not diagnosed on ultrasound scan but confirmed at delivery (n = 2) |
|---|---|---|---|---|
| Mode of delivery, n (%) | ||||
| Cesarean | 47 (100) | 13 (100) | 2 (50.0) | 2 (100) |
| Nonemergent | 40 (85.1) | 9 (69.2) | 1 (25.0) | 1 (50.0) |
| Emergent | 7 (14.9) | 2 (15.4) | 0 | 1 (50.0) |
| Vaginal | 0 | 2 (15.4) | 1 (25.0) | 0 |
| Gestation age at delivery by delivery mode, wk a | ||||
| At delivery | 34.69 ± 1.762 | 34.27 ± 1.988 | 37.04 ± 1.553 | 35.86 ± 0.404 |
| Nonemergent cesarean delivery and vaginal delivery | 34.68 ± 1.851 (n = 40) | 34.78 ± 1.474 (n = 11) | 37.04 ± 1.553 (n = 4) | 35.57 (n = 1) |
| Emergent cesarean delivery | 34.76 ± 1.236 (n = 7) | 31.50 ± 2.727 (n = 2) | 0 | 36.14 (n = 1) |
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
