Ultrasound examination has a central role, and a complex one. The many possible etiologies of FGR dictate a broad application of ultrasound principals in formulating a differential diagnosis. The many facets of fetal and maternal disease mean that multiple modalities must be used to form a comprehensive assessment of the FGR pregnancy. So, ultrasound evaluation is not simply that of quantifying fetal size and proportions, but is aimed at differentiating 3 basic clinical groups.
The constitutionally small fetus. These fetuses usually do not require serial evaluation once the diagnostic sequence is completed.
Fetal anomalies, including chromosomal, nonaneuploid, acquired (eg, viral infection). Prognosis is determined by the underlying problem, invasive testing is frequently needed, detailed family counseling is mandated, and decisions must incorporate the potential for a very poor prognosis.
Placenta-based FGR. This is the focus of ultrasound monitoring and FGR management, and must include gray-scale ultrasound, behavioral studies, and Doppler evaluation in the context of a multidisciplinary perinatal management strategy.
Biometry cannot serve as the sole diagnostic modality—simply demonstrating smaller than average measurements exaggerates the iatrogenic complications resulting from unnecessary preterm delivery.42 Similarly, relying on a single biometric parameter also induces unnecessary interference, and may miss major contributors to the true pathology of FGR. For example, the historic use of the biparietal diameter (BPD) to describe patterns of FGR resulted in unnecessary interference for a number of fetuses who had small heads as a result of aneuploidy or other complications, and ignored many fetuses who had normal head growth but marked reduction in other growth parameters.43 Current estimation of fetal weight involves presets within the ultrasound instrument, estimating fetal weight on a background of local growth curves using multiple parameters.44 The sonographer must be aware of the potential issues surrounding each individual measurement, while emphasizing that composite fetal weight estimation is the rule (Figure 9-4).
Ultrasound parameters of fetal growth estimation. Abdominal circumference (center) is the primary determinant for all equations. Clockwise from upper left: head circumference, femur length, humerus length, and transcerebellar diameter impact the equations in descending order.
There is broad variation in fetal head size, and in the relationship to gestational age, which becomes broader as the third trimester progresses. The majority of FGR fetuses have BPD measurements within the normal range, and other factors fairly common in the critical 28- to 32-week gestational time frame, associated with interference with BPD accuracy, include oligohydramnios, breech presentation, other malpresentation, and slight decline in head size compared with marked decline in abdominal circumference percentiles.45
Although somewhat less variable, head circumference (HC) has limitations that are similar to BPD—most asymmetric FGR (head circumference relatively normal, abdominal circumference reduced) will have late emergence of any changes in HC trend. An ancillary head measurement, the transcerebellar diameter, remains relatively stable in its relationship to gestational age, independent of FGR.46 The diagnostic strength of this measurement when dating is not known remains arguable—its clinical role is one of confirmation rather than proving gestational age.47
This is the most significant single factor in prediction of fetal weight and in evaluation of FGR.48 The methodology of abdominal circumference (AC) measurement is technically straightforward, but the parameters are exacting. In all multiple-parameter equations used to estimate fetal weight from ultrasound measurements, the AC has the most statistical influence. The positive predictive values of ultrasound measurements are enhanced by further parameters, but the AC has the most important role of all biometry. When the 10th percentile is used as the defining criterion for FGR, abdominal circumference alone has a higher sensitivity than multiformat ultrasound prediction of fetal weight.49 It is important to remember that the FGR fetus has a smaller liver, and therefore smaller abdominal circumference, due to the pathophysiology of disease and not related to gestational age. So, when dating is uncertain, abdominal circumference should not be used in the parameters chosen to estimate gestational age.
This improves detection of the fetus with asymmetric FGR, while the ratio remains normal in symmetric FGR, where both HC and AC are suppressed.50 Fetuses with nonplacental sources for FGR tend to concentrate in the symmetric group, but the specificity of this observation is not high enough for clinical diagnosis.51 The primary role for the HC/AC ratio is to alert managing caregivers to the onset of differential growth, as the ratio diverges earlier than the absolute measurements fall below the 10th percentile.
Femoral measurements, or measurements of the femur and humerus, add modestly to the sensitivity and specificity of estimated fetal weight calculations in predicting FGR.52 These measurements are more susceptible to asymmetric influences as pregnancy progresses and may improve the value of estimating fetal weight in the third trimester. The femur length (FL)/AC ratio is relatively static in normally grown fetuses for the entire second half of pregnancy. When the FL/AC ratio exceeds 23.5, regardless of gestational age or overall size of the fetus, FGR should be suspected.53
The ratios alluded to earlier reflect relative changes in proportion based on abdominal circumference being small in FGR. The ponderal index, used in neonatal medicine with superior prediction of asphyxia, acidosis, hypoglycemia, and impacts of intrauterine malnutrition, uses neonatal weight and length to estimate appropriateness of size.54 Direct translation into fetal measurements is difficult, because overall fetal length is difficult to evaluate directly and, as a function of femoral and other skeletal measurements, may amplify measurement errors. The intrauterine ponderal index has not gained broad acceptance, although in experienced hands it may improve the prediction of adverse outcomes in small fetuses.55 Similarly, subcutaneous and intracavitary fat layer measurements have had limited application.56 Reduced fetal fat deposition is a common feature of FGR, and fat deposition may be reduced more dramatically than other body components.57 Just as fat layer measurements in macrosomic fetuses may produce a relative impression of abnormal growth, absent subcutaneous or prerenal fat may approximate FGR. However, precise measurements vary highly between fetuses and are usually most applicable late in gestation, when relative FGR is more directly managed.
A number of calculations have been proposed, utilizing the individual components above. Those with most reliability use abdominal circumference as the primary influential factor, with small adjustments for skeletal growth, head size, and gestational age.44,51,58 Further specificity may be gained by correcting for fetal gender. Even more precise estimates of expected fetal weight may be discerned by using growth potential, the Gardosi approach described earlier.11 This estimate of fetal weight is less dependent on population growth curves and more influenced by maternal factors and fetal factors, including direct measurements and fetal gender, and calculating birth weight percentiles.59 Expected fetal growth can be estimated for full-term and for interval measurements during the pregnancy, and differences from this projection and the measured value can be determined.60 Although this has been applied successfully in small populations at risk for growth abnormalities where maternal characteristics do not conform to the surrounding "normal" population—for example, women with human immunodeficiency virus (HIV), women with chronic malnutrition, or specific disease states—overall application to general populations has not been shown to be superior.
The experienced ultrasonographer will have adequate data for formulating reference curves based on local normal populations, and be able to choose from the many programs offered with each ultrasound instrument to define that which best suits that population (Figure 9-5).
Serial growth demonstrating asymmetric FGR. Deviation from normal growth is first suggested by the elevation of the HC/AC ratio at 26 weeks. Head growth is relatively preserved, but both head and skeletal growth rates decline modestly, so the BPD/FL ratio remains stable. Estimated fetal weight is calculated from an equation including head circumference, abdominal circumference, and femur length. As FGR worsened, Dopplers progressed and, finally, amniotic fluid volume fell. The baby did not tolerate induction of labor and was delivered at 321/2 weeks by cesarean section, weighing 1250 g, 3rd percentile.
Measurement Definition of FGR
Several definitions have been used in clinical studies and can be related to adverse outcome, neurodevelopmental statistics, and long-term well-being. These include the following:
Estimated fetal weight (EFW) less than the 10th percentile for gestational age (±gender correction)
Abdominal circumference less than the 5th percentile for gestational age
Serial change in abdominal circumference less than 14 mm over 2 weeks
Use of a statistical threshold such as the 10th percentile is more a screening definition than it is based on physiologic principals. Of course, there are many 8th and 9th percentile fetuses that are normal, will have no adverse outcomes, and do not represent part of the spectrum of placental insufficiency. However, in clinical practice, setting the EFW at lower values will miss many fetuses with substantial nutritional and other deprivation that are at risk for intrauterine and newborn complications due to placental abnormalities. Estimating fetal weight has clinical variability, a large false-positive rate (about 70% of fetuses below the 10th percentile are simply small fetuses), and many will be shown to have abnormalities consistent with the congenital or fetal factors mentioned above. In general, however, the bottom 10% to 15% of fetuses defined by composite estimation of fetal weight account for a very high percentage of perinatal morbidity and mortality.61 Only the interaction with prematurity carries more significance in determining outcome, and in mandating high-risk strategies for monitoring and management. The lack of precision of these criteria emphasizes the importance of serial measurement. In some clinical situations, the recommended interval of 3 weeks between measurements (predicated on reducing the false-positive impression of stagnant growth) may be too long—the fetus with grossly abnormal placental and systemic circulation will be monitored at least weekly, and fetal growth estimates paired with monitoring intervals are appropriate. However, once growth parameters suggest significant FGR, the value of those measurements wanes. Indeed, moderate or severe FGR being diagnosed, functional evaluation of the fetus takes over, and further measurements have little significance in terms of management.