Preeclampsia is a gestation-specific syndrome that affects 2% to 8% of all pregnancies.¹ It is characterized by new onset hypertension and proteinuria beginning at or after 20 weeks’ gestation.² The incidence among nulliparous women is not only higher but has also increased in the United States, from approximately 12% to 28% over a period of 17 years.^3,4 Possible reasons for this increase include a trend toward higher BMI and maternal age at the time of conception and a higher number of pregnancies conceived using assisted reproductive technology (ART).⁵ Despite tremendous progress in obstetrical care during the 20th and early 21st centuries, preeclampsia and eclampsia continue to claim maternal and fetal lives and contribute significantly to neonatal morbidity and mortality.^6-8 This is largely due to a lack of effective strategies to prevent and treat the disease. The only known cure is to remove the placenta, which implies early delivery and prematurity in a substantial proportion of cases.²

The etiology of preeclampsia is multifactorial. However, impaired placentation is considered to play a key role in the development of the disease, particularly early onset preeclampsia, ie, preeclampsia requiring delivery before 34 weeks.² It is not clear whether late onset preeclampsia has a different pathogenic mechanism or represents a different form of the same disease.^2,9 Although late onset preeclampsia represents the majority of the cases (approximately 78%-88%),^10,11 early onset preeclampsia is associated with a high frequency of histologic placental lesions consistent with maternal underperfusion¹² and a significantly higher risk of perinatal complications (ie, small-for-gestational age [SGA] status, fetal and neonatal death, and combined perinatal death and morbidity).¹¹

Low-dose aspirin has been proposed as a prophylactic agent to prevent preeclampsia related to impaired placentation. A meta-analysis of 31 trials showed that the administration of low-dose aspirin to high-risk patients is associated with a reduction of approximately 10% in the rate of preeclampsia (relative risk [RR] = 0.90, 95% confidence interval [CI] = 0.84-0.97).¹³ A meta-analysis published in 2010, with analysis stratified according to the gestational age at which low-dose aspirin prophylaxis was initiated, showed a 50% reduction in the prevalence of preeclampsia when prophylaxis started at 16 weeks or earlier (RR = 0.47, 95% CI = 0.34-0.65), and a larger benefit observed when the outcome was severe preeclampsia (RR = 0.09, 95% CI = 0.02-0.37).¹⁴ A more recent meta-analysis including the results of 45 randomized clinical trials with a total of 20909 pregnancies randomized to either low-dose aspirin or placebo, showed a significant reduction in the rates of preeclampsia (RR = 0.57; 95% CI = 0.43-0.75; p <0.001; R2 = 44%; p = 0.036), severe preeclampsia (RR = 0.47; 95% CI = 0.26-0.83; p = 0.009; R2 = 100%; p = 0.008) and fetal growth restriction (FGR) (RR = 0.56; 95% CI = 0.44-0.70; p <0.001; R2 = 100%; p = 0.044). Higher doses of aspirin (ie, 100 mg compared to 60 mg) were associated with more significant reduction in the rate of preeclampsia.¹⁵

Identification of women at risk for preeclampsia is a worthwhile goal of prenatal care. Uterine artery Doppler has been proposed as early as 1983 as a screening test for the condition.¹⁶ Population screening would be justified if: (1) the test (or combination of tests) has high sensitivity and a reasonable false-positive rate, (2) the test is reproducible (ie, can be replicated in different clinical environments with the same level of accuracy as in the research environments where the test was developed), and (3) there is an effective preventive strategy or treatment for the disorder being screened for. In addition, the test should be relatively low cost, to be applied to a large population. In this chapter, we will review the anatomy of the uteroplacental circulation, the physiological changes that this circulation undergoes as a result of pregnancy, and the technique for Doppler assessment of the uterine arteries as well as screening strategies to predict preeclampsia and small-for-gestational age (SGA) fetuses in the first and second trimesters of pregnancy.

The uterine arteries are branches of the internal iliac arteries. Upon reaching the isthmic portion of the uterus, the uterine artery gives off a small cervical branch before ascending along the lateral uterine wall to anastomose with the ipsilateral ovarian artery in the cornual region on each side. Arcuate arteries are branches of the uterine arteries that run along the anterior and posterior surfaces of the uterus in the transverse plane. Approximately 100 radial arteries branch off the arcuate arteries at an approximately 90-degree angle to penetrate the myometrium towards the decidua. Upon reaching the decidua, the radial arteries branch into spiral arteries, which supply the intervillous space with oxygenated maternal blood.

The spiral arteries have a well-developed muscular layer that is responsible for the low capacitance and high-resistance to blood flow of the uterine circulation in the nonpregnant state. Physiologic transformation of the spiral arteries is a term that describes invasion of the spiral arteries by trophoblast during the first (primary invasion soon after implantation) and early second trimesters (secondary invasion, around 12-16 weeks) of pregnancy leading to replacement of the endothelium and actual disappearance of the vessels muscular layers, specifically during secondary invasion. Thus, in normal pregnancies, the uteroplacental circulation is transformed from low capacitance/high resistance to a high capacitance/low resistance vascular bed that is also unresponsive to maternal vasomotor control.^17-23

Failure of the previously described physiologic transformation of the spiral arteries is associated with several pregnancy disorders, including preeclampsia, fetal growth restriction (FGR), late spontaneous abortion, abruptio placentae, preterm labor, and premature rupture of the membranes.^24-35 The strongest association reported to date is with preeclampsia, particularly early onset preeclampsia, which is the form associated with the higher incidence of adverse outcomes.^36,37

Doppler velocimetry of the uterine arteries assesses the summation of resistances encountered along the more distal vessels of the vascular bed, ie, the spiral arteries. Thus, failure of physiologic transformation of the spiral arteries can be inferred by demonstrating increased resistance to blood flow in the uterine arteries.

Technique can be found in the ISUOG guidelines on the use of Doppler ultrasonography in obstetrics.³⁸

First Trimester

Uterine artery Doppler velocimetry may be performed using transabdominal or transvaginal ultrasonography. Regardless of the method used, the examiner identifies the uterine arteries by color Doppler on a sagittal section on each side of the cervix and uterus at the level of the internal os prior to obtaining a waveform. Once the vessel is identified it should be magnified before interrogating it with pulsed Doppler. Sampling gate should be set at 2 mm to cover the whole vessel. The uterine arteries appear more tortuous during the first trimester when compared to their appearance during the second trimester, and the examiner should strive to sample the vessel as it courses upwards towards the transducer, keeping the angle of insonation at less than 30 degrees. The PI of 3 similar consecutive waveforms should be measured on each side and the mean PI calculated. By using this technique, the peak systolic velocity of the uterine artery should be >60 cm/s, avoiding erroneous sampling of one of the arcuate arteries.^39,40 Figure 11-1 shows a uterine artery Doppler waveform obtained at 12 weeks of gestational age. Please note that an early diastolic indentation, known as a “notch” (arrow), is seen in approximately 45% of the patients examined during the first trimester.^41,42 Although this indicates elevated resistance to blood flow during the late second trimester, the presence of a diastolic notch is a normal phenomenon during the first trimester. Doppler insonation of the uterine arteries does not impair risk to the fetus.⁴³

Second Trimester

During the second trimester, the approach to obtain a uterine artery Doppler waveform changes as the uterus enlarges and the uterine arteries are identified in a more lateral position in the lower abdomen. In order to identify the uterine arteries on each side, the examiner has to position the transducer slightly more laterally than the uterine wall and then orient the transducer face medially and inferiorly. The uterine artery is seen by color Doppler as a large vessel coursing obliquely towards the transducer at a point where it appears to cross over the external iliac artery and vein (Figure 11-2). This is an apparent crossing caused by the angle at which the uterine arteries are insonated as, in reality, the external iliac artery and vein are more superficial than the uterine artery. A normal second trimester Doppler velocity waveform is shown in Figure 11-3.

Resistance Indices

The most basic hemodynamic parameters assessed by Doppler velocimetry are blood flow direction and velocity. Accurate estimation of velocity requires that the angle of insonation be known and kept below 20 degrees at all times. As the cosine of the angle of insonation is a component of the numerator of the Doppler equation, insonation at 30-, 45-, and 60-degree angles underestimate velocity by 13%, 29%, and 50%, respectively (provided that an angle corrector is not used). Fortunately, resistance to blood flow in a circulation can be assessed with angle independent indices, namely the pulsatility index (PI), resistance index (RI), and systolic-to-diastolic ratio (commonly known as S/D or A/B ratio). Among these, the PI is the index that best correlates with actual resistance to blood flow. Support for this affirmation originates from a study of 4 pregnant ewes examined between 16 and 17 weeks of gestation.⁴⁴ Blood flow volume and resistance to blood flow were monitored invasively, and Doppler velocimetry measurements were obtained while the uterine circulation was embolized with microparticles (Gelfoam) with a concentration of 0 to 35 mg over a period of 3 hours. The only parameter that had a positive linear correlation with actual resistance to blood flow was the PI. The correlation between both RI and S/D ratio was nonlinear, and therefore these indices only began to raise when actual resistance to blood flow reached a nadir of 200 mm Hg. The important message from this study is that the RI and S/D ratios may not have enough sensitivity to detect mild increases in peripheral vascular resistance when compared to the PI.

What Is Normal?

A diagnostic test is only useful once one knows what is normal. Determining what is normal requires applying the test to large populations of normal patients and conducting regression analysis to determine measures of central distribution (means or medians) and dispersion (standard deviations or percentiles). Studies of uterine artery Doppler velocimetry have shown that: (1) resistance to blood flow (and therefore the PI, RI, and S/D ratios) decreases as pregnancy advances from the first to the third trimester; (2) approximately 45% of the waveforms obtained in normal pregnancies during the first trimester have a diastolic notch^41,42; (3) the diastolic notch disappears in the majority of pregnancies in the second trimester (after 22-23 weeks), being present in only approximately 5% of normal pregnancies at that time; and (4) resistance to blood flow is lower in the uterine artery ipsilateral to the site of placental implantation.⁴⁵ Given that resistance to blood flow may be higher in the uterine artery contralateral to the site of placental implantation, researchers usually report either the mean PI (mean UtA PI) between the two uterine arteries or results of the uterine artery with the lowest PI (lowest UtA PI).^10,46,47 The 95th percentile for the mean uterine artery PI in the first trimester has been estimated at 2.35 according to a study of 3324 women examined between 11 and 14 weeks by Martin et al.⁴⁶ In another study conducted in Australia, Ridding et al⁴⁸ reported a mean PI of 1.65 ± 0.41 (1SD) as the normal values between 11 and 14 weeks. The authors also reported the normal values for the uterine artery with the lowest PI as a mean of 1.44 ± 0.38 (1SD). For examinations performed during the second trimester, the 95th percentile for the mean PI of the uterine arteries examined between 22 and 24 weeks has been estimated as 1.61 (median 1.04) among 16806 patients examined by Papageorghiou et al⁴⁹ Several other nomograms have been published describing normal values for examinations performed during these and other gestational ages.^41,48,50-57

The presence (or better stated, the persistence) of bilateral diastolic notches after 22 to 23 weeks of gestation is considered abnormal and is strongly associated with an increased risk for preeclampsia.^55,58-61 Images illustrating abnormal second trimester uterine artery Doppler waveforms are shown in Figure 11-4.

Reproducibility

Consistency in measurement is a requirement for the successful performance of any diagnostic test.⁶² Previous research has shown that there is potential for extensive variability in uterine artery Doppler measurements depending on the site of measurement and other technical factors.⁶³ A prospective study of uterine artery Doppler measurements obtained by 12 different operators between 11 and 13 weeks and 6 days of gestation, showed a higher screen-positive rate to predict preeclampsia for inexperienced compared to experienced examiners (10% vs 2.7%, p <0.0001). Moreover, examiners who received regular feedback regarding audit of their measurements every 4 weeks had significantly lower median multiples of the median (MoM) uterine artery PI (L-PI) measurements (mean log10 L-PI -0.0053 vs 0.0044, p = 0.0035) and a lower rate of screen-positive results when compared to examiners who did not receive any feedback (4.3% vs 6.0%, p <0.012).⁶² This underscores the importance of training and an audit system in place to ensure accuracy of the test before it can be implemented clinically.

Uterine artery Doppler velocimetry has been studied mostly as a screening test to identify women at high risk to develop preeclampsia and, to a lesser extent, FGR. Outside of screening, the test may be used in the evaluation of a fetus incidentally found to measure smaller than expected for gestational age, as those with elevated uterine artery Doppler resistance to blood flow are more likely to be small due to placental insufficiency. Papageorghiou et al, for example, have shown that a substantial proportion (approximately 40%) of fetuses identified in the second trimester as having a short femur actually have early FGR and found that the uterine artery Doppler was abnormal in 90% of these cases at the time of presentation.⁶⁴

Screening for Preeclampsia

At the time of this writing, guidelines issued by several professional societies recommend risk assessment for preeclampsia based on maternal historical factors at the time of the first prenatal visit and initiation of low-dose aspirin as a prophylactic agent (75-162 mg) from 12 weeks until delivery for pregnancies considered at high-risk. (Table 11-1).^65-69

Screening strategies to refine the identification of pregnancies at risk for preeclampsia have been extensively investigated and include, besides historical factors listed on Table 11-1, maternal blood pressure measurements, uterine artery Doppler velocimetry, maternal serum biomarkers, and metabolomics.

Screening Based on Maternal History

In a study of 8366 women attending their first routine prenatal visit, screening for preeclampsia using maternal demographic characteristics was only useful when risk factors were combined into an algorithm derived by multivariate analysis. Using such an algorithm, 36%, 33%, and 29% of the cases of preeclampsia manifesting before 34 weeks, 37 weeks, and 42 weeks respectively, were correctly identified for a fixed 5% false-positive rate (FPR). For a fixed 10% FPR, correct prediction of preeclampsia <34 weeks, <37 weeks, and <42 weeks increased to 51%, 43%, and 40%, respectively. Predictors of early preeclampsia were black race, chronic hypertension, prior preeclampsia, and use of ovulation drugs. Increased maternal age, high body mass index (BMI), and family history or prior history of preeclampsia predicted late preeclampsia.⁷⁰

A recent meta-analysis of 92 publications comprising 25,356,688 pregnancies in 27 countries investigated the association between several clinical risk factors and the development of preeclampsia in the index pregnancy.⁷¹ The results of this meta-analysis, which are summarized on Table 11-2, showed that the strongest clinical risk factors for the development of preeclampsia were history of antiphospholipid syndrome, chronic hypertension, prior history of preeclampsia, pregestational diabetes, prior placental abruption, multifetal pregnancy, prepregnancy BMI >30 and also >25, ART, chronic kidney disease, prior stillbirth, maternal age >40 and to a lesser extent >35, nulliparity, and systemic lupus erythematosus (SLE).

Screening Based on Maternal Blood Pressure

An increase in blood pressure can be observed in women at higher risk to develop preeclampsia as early as the first trimester. A systematic review of studies evaluating the usefulness of maternal blood pressure as a risk factor for the subsequent development of preeclampsia showed that mean arterial blood pressure (MAP) performs significantly better than systolic or diastolic blood pressure.⁷² Measurement instrumentation and technique are important to increase the accuracy of blood pressure measurements. Thus, MAP should be measured using validated automated devices. In addition, women should be sitting on a chair with good back support, the legs should be uncrossed, and two simultaneous measurements should be taken from each arm, which are to be supported at the level of the heart. An average of four measurements should be used.⁷³ The sensitivities to predict preeclampsia less than 34 weeks, 37 weeks, and 42 weeks with MAP measurements added to maternal historical risk factors have been estimated at 58%, 44%, and 37%, respectively, for a fixed 5% FPR. For a 10% FPR, detection rates increased to 73%, 59%, and 54%, respectively.

Screening Based on Serum Biomarkers

A wide array of serum biomarkers have been tested to predict preeclampsia, the best known of which are plasma-associated pregnancy protein-A (PAPP-A), placental growth factor (PlGF), soluble Fms-like tyrosine kinase-1 (sFlt-1), placental protein-13 (PP13), inhibin-A, activin-A, pentraxin-3 (PTX-3), P-selectin, and disintegrin and metalloproteinase 12 (ADAM12). sEng, inhibin-A, activin-A, PTX3, P-selectin, and ADAM-12 are increased, and PAPP-A, PlGF, and PP13 are decreased compared to controls in patients who develop preeclampsia.^74,75 However, no single biomarker has been shown, in isolation, to be sensitive and specific enough to be used as a screening test to predict preeclampsia.^75-79 Similarly, a combination of maternal serum biomarkers, maternal historical risk factors, and blood pressure measurements has proven ineffective as a screening strategy to predict preeclampsia in low risk nulliparous women (sensitivity 46.1%, 95% CI 38.3-54.0; specificity 80%).⁸⁰

Metabolomics

Several groups are currently investigating the role of metabolomics to predict preeclampsia. Kenny et al⁸¹ reported in 2008 that uric acid, 2-oxoglutarate, glutamate, and alanine were discriminatory biomarkers that could be used to detect preeclampsia in maternal plasma. In a subsequent discovery-validation study, the same group reported that a multivariate model combining 14 metabolites from maternal plasma obtained at 15 ± 1 weeks’ gestation was associated with an odds ratio of 23 (95% CI = 7-73) for the development of preeclampsia.⁸²