Objective
The frequency of fetal anomalies in women with pregestational diabetes correlates with their glycemic control. This study aimed to assess the predictive performance of first-trimester fetal nuchal translucency (NT), ductus venosus (DV) Doppler, and hemoglobin A1c (HbA1c) to predict fetal anomalies in women with pregestational diabetes.
Study Design
This was a prospective observational study of patients undergoing first-trimester NT with DV Doppler. Screening performance was tested for first-trimester parameters to detect fetal anomalies.
Results
Of 293 patients, 17 had fetal anomalies (11 cardiac, 7 major, 3 multisystem). All anomalous fetuses were suspected prenatally. One had NT >95th centile, 2 had reversed DV a-wave, and 13 had HbA1c >7.0%. The HbA1c was the primary determinant of anomalies (r 2 , 0.15; P < .001) and >8.35% was the optimal cutoff for prediction of anomalies with an area under the curve of 0.72 (95% confidence interval, 0.57–0.88). Therefore, first-trimester prediction of anomalies was best in women with increased NT or HbA1c >8.3% (sensitivity 70.6%, specificity 77.4%, positive predictive value 16.2%, negative predictive value 97.7%, P < .001).
Conclusion
In women with pregestational diabetes and poor glycemic control, an increased NT increases risks for major fetal anomalies. Second-trimester follow-up is required to achieve accurate prenatal diagnosis.
Women with pregestational diabetes mellitus are at substantially increased risk for having a fetus with congenital anomalies. This risk is related to periconceptional glycemic control as evidenced by the higher frequency of fetal anomalies in women with raised glycosylated hemoglobin (HbA1c) levels at the beginning of pregnancy. Fetal anomalies predominantly involve cardiac, musculoskeletal, urogenital, and central nervous systems and of these cardiovascular anomalies comprise the highest proportion and are the greatest contributor to perinatal mortality. In recognition of the frequency and pattern of anomalies, typical prenatal screening for women with pregestational diabetes consists of a targeted second-trimester ultrasound followed by fetal echocardiography. However, since the risk for fetal anomalies in pregestational diabetes can already be estimated in early pregnancy by the HbA1c level these women may benefit from potential early detection of fetal anomalies in the first trimester. Such early detection can offer the advantage of early consideration of management options, early risk stratification, and, if pregnancy termination is desired, the advantage of increased privacy and decreased cost, morbidity, and mortality.
First-trimester ultrasound measurement of the fetal nuchal translucency (NT) thickness and integration of this measurement with additional parameters not only identifies risk for aneuploidy with greater accuracy than second-trimester screening but also predicts fetal anomalies. In addition to fetal aneuploidy, an increased NT and abnormal ductus venosus (DV) flow velocity waveform can independently predict cardiovascular and renal anomalies and adverse perinatal outcome. As many of the anomalies observed in pregestational diabetes are also associated with an increased NT and since the dimension of the NT appears to be independent of the HbA1c or the degree of glycemic control, first-trimester screening using NT thickness and DV Doppler may lead to their early identification.
Accordingly, it was the aim of this study to test the hypothesis that either an increased first-trimester NT or abnormal DV Doppler will detect fetal anomalies in women with pregestational diabetes, especially in association with an elevated HbA1c.
Materials and Methods
This is a prospective, observational study of women with pregestational diabetes presenting for first-trimester screening at the University of Maryland Center for Advanced Fetal Care and the Division of Maternal-Fetal Medicine, Eastern Virginia Medical School, from April 2003 through June 2009. The study protocol was approved by the institutional review boards of both institutions. Inclusion criteria were preexisting diabetes mellitus requiring medical therapy with confirmed viable intrauterine pregnancy in women who presented for first-trimester screening. Maternal demographics, pregnancy/diabetes history, and HbA1c were obtained as a part of routine clinical practice. Patients agreed to follow-up evaluation as part of the study.
Women underwent a standardized first-trimester ultrasound examination between 11+0 and 13+6 weeks’ gestation (crown-rump length [CRL], 45-84 mm). The gestational age was calculated from the last menstrual period based on the Naegele rule. If there was a >1 week discrepancy between the calculated gestational age and the CRL, the dates were adjusted to the latter. In addition to the CRL measurement, the examination included measurement of the NT; documentation of the head, brain, stomach, abdomen, bladder, and extremities; and 4-chamber heart view. The DV waveform was assessed by pulsed wave Doppler as previously described and the length of the examination was limited to 2 minutes with power settings adjusted to minimize fetal exposure. When the typical DV waveform was obtained, the image was frozen, the pulsatility index was measured, and the direction of blood flow during atrial systole was visually assessed as antegrade, absent, or retrograde. NT and DV Doppler measurements were performed by certified sonographers according to Fetal Medicine Foundation or NT Quality Review Program. Either transabdominal or transvaginal approach was used and time to complete the examination was at the discretion of the examiner to optimize visualization of all structures.
Following the first-trimester ultrasound, women with continuing pregnancies underwent second-trimester targeted ultrasound at 18-20 weeks’ gestation and fetal echocardiography at 22-24 weeks’ gestation according to recommendations of the International Society of Ultrasound in Obstetrics and Gynecology. Subsequent ultrasounds for growth and fetal assessment were performed as clinically indicated.
At the first-trimester sonogram, fetal anomalies were only reported if the diagnosis was certain. Patients with suspected anomalies were informed of the findings with a caveat that a follow-up sonogram would focus on a more complete evaluation of the suspected anomaly. For patients where a prenatal diagnosis was possible with a reasonable amount of certainty, fetal anomalies were documented and classified according to organ system. An anomaly was defined as a malformation resulting from abnormal development. Major fetal anomalies were defined as malformations that are lethal, cause serious impairment, or require intervention. Those that were present but anticipated to have minimal functional implications were defined as minor. Additional ultrasound markers and findings were recorded if observed. Patients were managed according to the standard of care regardless of study participation.
Pregnancy outcome and delivery details were obtained from patient records or obstetric providers. An attending pediatrician performed the neonatal evaluation prior to hospital discharge. Anomalies were confirmed and any additional findings were documented.
NT thickness >95th centile for CRL and reversal of the a-wave in the DV were evaluated as first-trimester predictors of fetal anomalies. According to the recommendations of the American Diabetes Association, HbA1c >7% defined poor glycemic control. The presence of either an abnormal NT or DV was tested as a predictor of any fetal anomaly, major anomalies, and specific organ system anomalies. The predictive accuracy was also evaluated stratified by an abnormal HbA1c. Fisher exact test and χ 2 test were used as appropriate for analysis of categorical variables. Mann Whitney U test was used for comparison of continuous variables. Receiver operating characteristic (ROC) curves were constructed to identify if cutoffs with superior first-trimester screening performance could be identified for continuous variables. Area under the ROC curves and 95% confidence intervals were compared and logistic regression was used to identify the primary predictors of fetal anomalies. A P value < .05 was considered statistically significant. Statistical software IBM SPSS Statistics 20 (SPSS Inc., Chicago, IL) was used for the analysis.
Results
From April 2003 through June 2009, 341 women consented for study participation. The first-trimester ultrasound revealed miscarriage in 3, and a fetal CRL outside the gestational age criteria in 23 women. In 10 women, NT or DV Doppler could not be obtained; 11 women were lost to follow-up and 1 patient withdrew participation leaving 293 patients for final analysis ( Table 1 ).
| Parameter | Data |
|---|---|
| Maternal characteristics | |
| Age, y | 30 (18–45) |
| Gravidity | 2 (1–17) |
| Parity | 1 (0–14) |
| Ethnicity, n (%) | |
| African American | 159 (53.6) |
| White | 122 (41.5) |
| Asian | 5 (1.7) |
| Hispanic | 3 (1) |
| Asian Indian | 1 (0.3) |
| Undocumented | 3 (1) |
| BMI (kg/m 2 ) | 34.6 (18.5–76.4) |
| BMI (kg/m 2 ) >30, n (%) a | 127 (68.6) |
| BMI (kg/m 2 ) >40, n (%) a | 48 (25.9) |
| First-trimester parameters | |
| HbA1c, g/dL | 7.0 (4.8–15.4) |
| GA, wk | 12.5 (11.1–14.4) |
| CRL, mm | 64.6 (42.6–86) |
| NT, mm | 1.4 (1.0–7.7) |
| DVPI | 1.07 (0.9–2.37) |
| Perinatal outcomes b | |
| GA at delivery (wk), median (interquartile range) | 37.6 (36.1–38.5) |
| Birthweight (g), median (interquartile range) | 3298 (2648–3725) |
| Live birth | 279 (95.2) |
| Spontaneous miscarriage | 7 (2.4) |
| Elective termination | 5 (1.7) |
| FDIU | 2 (0.7) |
| Cesarean delivery | 180 (61.4) |
| Apgar <5 at 5 min | 1 (0.4) |
| NICU admission | 73 (24.9) |
a Percent of 185 patients with recorded BMI;
Seventeen women had an anomalous fetus, of which 7 cases were major anomalies and 10 were minor anomalies. Major anomalies were diagnosed at each examination with most of them being identified in the first trimester (n = 2, 28.6%) or at the time of fetal echocardiogram (n = 3, 42.9%). The second-trimester detailed anatomy scan and interval growth scan each identified 1 fetus with a major anomaly. The second-trimester ultrasound examination was most important for identifying minor anomalies (n = 6, 60%) but additional minor anomalies were detected at fetal echocardiogram (n = 3, 30%) and at the interval growth scan (n = 1, 10%) As a single organ system cardiac anomalies were most frequent (n = 11, 65%) ( Table 2 ). None of the cases with isolated ventricular septal defects required surgical intervention in the immediate postnatal period. Limited visualization of ≥1 organ systems occurred in 13 cases (4.4%) at the second-trimester anatomy scan.
| Case | GA, wk | BMI, kg/m 2 | CRL, mm | HbA1c, % | NT, mm | DV PI | DV a-wave | GA Dx, wk | Outcome | Anomaly |
|---|---|---|---|---|---|---|---|---|---|---|
| 1 | 11.6 | 49.7 | 55 | 8.7 | 1 | 1 | Positive | 22 | Live birth | TOF |
| 2 | 13.2 | 41 | 70.7 | 7.1 | 1.5 | 1.14 | Positive | 23.4 | Live birth | TOF, holoprosencephaly, cleft lip/palate |
| 3 | 12.8 | 76.4 | 73 | 8.6 | 2.6 | 1 | Positive | 29.3 | Live birth | TOF, AV canal defect |
| 4 | 12.1 | 27.5 | 56.6 | 15.4 | 1.6 | 1.27 | Positive | 25.1 | Live birth | ASD/VSD, pulmonary stenosis, hypospadias a |
| 5 | 12.3 | 39.6 | 66.1 | na | 1.6 | na | Reversed | 12.3 | TOP | Encephalocele |
| 6 | 13.6 | 28.9 | 73 | 12.1 | 1.3 | 1.12 | Positive | 13.6 | TOP | Omphalocele, VACTERL association, cerebellar hypoplasia b |
| 7 | 12 | 29.3 | 56.5 | 9.5 | 2.7 | 1.06 | Positive | 16 | TOP | Fused thalami, VSD, ventriculomegaly, 2VC, megacystis |
| 8 | 12.5 | 26.7 | 65.1 | 12.1 | 1.8 | 0.95 | Positive | 26.4 | Live birth | Bicuspid aortic valve, mild RVH |
| 9 | 13.1 | 31.6 | 69.9 | 9.6 | 2.4 | na | Positive | 19.1 | Live birth | VSD |
| 10 | 14.2 | 36 | 83 | 10.7 | 1.5 | 1.06 | Positive | 19.6 | Live birth | VSD |
| 11 | 12.3 | 23.9 | 63.4 | 5.8 | 1.2 | 0.9 | Positive | 19 | Live birth | VSD |
| 12 | 12.4 | na | 59.6 | 5.5 | 1.4 | 0.83 | Positive | 21.5 | Live birth | VSD, pericardial effusion |
| 13 | 13.6 | 33 | 71 | 10.4 | 1.2 | 2.3 | Reversed | 24.1 | FDIU | 2VC, echogenic bowel |
| 14 | 12.6 | na | 51.8 | 9.2 | 1.3 | 1.04 | Positive | 19 | Live birth | 2VC, placental vascular mass |
| 15 | 13.5 | 32.5 | 73.7 | 10.8 | 1.2 | 0.99 | Positive | 19.5 | Live birth | 2VC, hydrocele |
| 16 | 12.4 | 55 | 59 | 8.4 | 1.4 | 0.62 | Positive | 18.6 | Live birth | Pulmonary stenosis, a liver calcifications |
| 17 | 11.5 | 24.7 | 53.2 | 5.7 | 1.6 | 0.6 | Positive | 24.3 | Live birth | Persistent right umbilical vein |
a Finding diagnosed postnatally;
b Omphalocele was diagnosed in first trimester, additional anomalies were diagnosed early in second trimester prior to elective termination of pregnancy.
Of pregnancies, 95% ended in live birth and 70% of these were at term. Two fetal demises occurred: 1 in the context of severe preeclampsia and 1 due to fetal growth restriction, however autopsy data were not available to identify the precise cause. Of 5 elective terminations, 2 were for major fetal anomalies, 1 was for previable preterm premature rupture of membranes, 1 was for a fetus with trisomy 21, and 1 was for an anomalous fetus with a high suspicion of trisomy 21 with unsuccessful karyotype culture (case 7) ( Table 1 ). There were no other documented cases of aneuploidy.
Baseline characteristics between pregnancies with and without anomalies were similar, except women with an anomalous fetus had higher first-trimester HbA1c. These pregnancies were also delivered at earlier gestational age and less likely to result in live birth. Neonates with anomalies had lower birthweight and were more likely to require neonatal intensive care unit admission ( Table 3 ).
| Characteristic | Anomaly (n = 17) | No anomaly (n = 276) | P value |
|---|---|---|---|
| Age, y | 31.1 ± 6.4 | 30.2 ± 5.8 | .508 |
| BMI, kg/m 2 | 37.2 ± 13.6 | 34.9 ± 8.8 | .938 |
| HbA1c, % | 9.4 ± 2.6 | 7.3 ± 1.9 | .001 |
| NT, mm | 1.6 ± 0.5 | 1.5 ± 0.6 | .425 |
| NT >95th percentile | 1 (6) | 4 (1) | .277 a |
| DV PI | 1.1 ± 0.4 | 1.1 ± 0.3 | .172 |
| DV reversed a-wave | 2 (12) | 6 (2) | .77 a |
| GA at delivery, wk | 33.5 ± 6.5 | 36.4 ± 4.2 | .009 |
| Birthweight, g | 2470 ± 1311 | 3183 ± 882 | .045 |
| Live birth | 13 (76) | 266 (96) | .005 a |
| Cesarean delivery | 8 (47) | 172 (62) | .189 |
| NICU admission | 9 (53) | 64 (23) | .021 |
a Fisher exact test; all others are χ 2 or Mann-Whitney U tests.
For pregnancies ending in live birth, all anomalies were confirmed postnatally and additional anomalies were diagnosed in 2 cases. In 1 fetus with a complex cardiac defect hypospadias was diagnosed neonatally (case 4). In another fetus with second-trimester ultrasound evidence of liver calcifications, pulmonary stenosis was diagnosed postnatally (case 16).
Of the 17 fetuses with anomalies, 1 had NT >95th centile, 2 had reversed DV a-wave, and 13 had HbA1c >7%. Neither NT >95th centile nor reversed DV a-wave alone was predictive for anomalies. Detection rates for NT, DV Doppler, HbA1c, and various combinations are shown in Table 4 . ROC curve analysis did not identify better predictive cutoffs for the NT thickness or the DV pulsatility index. For the HbA1c, ROC curve analysis identified a value >8.35% as the optimal first-trimester screening cutoff for prediction of all anomalies (75% sensitivity; 76% specificity; area under the curve, 0.724; 95% confidence interval, 0.571–0.878) ( Figure ).
| Parameter | n/n | Sensitivity, % | Specificity, % | PPV, % | NPV, % | FPR, % | LR | Area under curve (95% CI) |
|---|---|---|---|---|---|---|---|---|
| NT >95th centile or DV reversed a-wave | ||||||||
| Any anomaly | 3/17 | 17.6 | 96.4 | 23.1 | 95 | 3.6 | 4.9 | 0.55 (0.40–0.71) |
| Major anomaly | 2/7 | 28.6 | 96.2 | 15.4 | 98.2 | 3.8 | 7.5 | 0.57 (0.32–0.83) |
| HbA1c >7% | ||||||||
| Any anomaly | 13/16 | 81.3 | 50.6 | 9.6 | 97.7 | 49.4 | 1.6 | 0.66 (0.54–0.78) |
| Major anomaly | 6/6 | 100 | 49.8 | 4.4 | 100 | 50.2 | 1 | 0.75 (0.63–0.87) |
| HbA1c >8.35% or NT >95th centile | ||||||||
| Any anomaly | 12/17 | 70.6 | 77.4 | 16.2 | 97.7 | 22.6 | 3.1 | 0.74 (0.61–0.87) |
| Major anomaly | 5/7 | 71.4 | 75.7 | 6.8 | 99.1 | 24.3 | 2.9 | 0.77 (0.60–0.95) |
| CV anomaly | 8/11 | 72.7 | 76.4 | 10.8 | 98.6 | 23.6 | 3.1 | 0.72 (0.57–0.88) |
| HbA1c >8.35% or NT >95th centile or DV reversed a-wave | ||||||||
| Any anomaly | 13/17 | 76.5 | 73.6 | 15.1 | 98.1 | 26.4 | 2.9 | 0.73 (0.6–0.86) |
| Major anomaly | 6/7 | 85.7 | 72 | 7 | 99.5 | 28 | 3.1 | 0.76 (0.59–0.94) |
| CV anomaly | 8/11 | 72.7 | 72.3 | 9.3 | 98.6 | 27.7 | 2.6 | 0.71 (0.56–0.87) |
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
