Integrated fetal testing in growth restriction: combining multivessel Doppler and biophysical parameters

The criteria for a successful screening test are well defined. The targeted condition must be an important health problem—at least 80% of the population should be at risk. The epidemiology, natural history and disease spectrum must be well delineated. There must be an identifiable early stage with a sufficiently long latent phase to allow intervention, which will change outcome. The employed tests should be simple, precise and applicable to the target population. The tests should be well validated with a known distribution of results and a suitable cut-off level to identify pre-disease. If a screening policy is to be finalized, there should be evidence from high-quality randomized controlled trials that the screening program is clinically, financially, socially and ethically acceptable and is effective in reducing morbidity or mortality. And there should be a plan for managing and monitoring the screening program and an agreed set of quality assurance standards. These criteria have yet to be fulfilled for any test of fetal well-being. Cardiotocography (CTG), Doppler and biophysical profile score (BPS) are the principal antenatal testing modalities. These have predominantly been compared in their ability to predict fetal compromise and optimize the timing of intervention1, 2. However, the prognostic and diagnostic information gained in each testing modality is in great part independent of each other. The question that I will address here is: might it be better to combine, or integrate these modalities3? We have no accurate estimate of the impact of IUGR since the effects of this condition extend from fetal life all the way into adulthood4. The interactions between maternal, placental and fetal factors in the regulation of growth and development are complex. While the impact of these factors on long- and short-term outcome is still under investigation, considerable insight into natural disease history in fetal life has been gathered. Abnormal villous development and/or trophoblast invasion compromises fetal nutrient and waste exchange and causes disturbed placental blood flow dynamics. This condition may manifest as overt abnormalities of blood flow in fetoplacental and uterine vessels, fetal growth restriction, abnormal fetal behavior, abnormal fetal heart rate patterns, a decline in amniotic fluid production, or a variable combination of the above. Once a discrepancy between placental supply and fetal demand is established, there are several possibilities. If placental disease precludes fetal growth and development, stillbirth occurs. If placental disease is less severe and survival is made possible through a series of adaptations, the price may be long-term dysfunction. If placental disease is very mild, growth and development may still be perceived to be normal but fall short of the genetically determined potential. Decompensation may occur at any time, triggered by aggravation of the maternal, placental or fetal condition, or by the stress of labor. Irrespective of intervention the fetus remains at risk for additional damage and stillbirth. A variety of cardiovascular and central nervous system (CNS) responses to placental insufficiency have been described. Evidence of abnormal placentation may be found in the uterine circulation as elevation of the Doppler resistance index or persistence of an early diastolic notch5. In the umbilical circulation, end-diastolic velocity may be normal, decreased, absent or reversed (A/REDV) proportional to the degree of villous vascular damage6. Elevated placental blood flow resistance favors redistribution of cardiac output towards vital organs7. This may be apparent through increased aortic blood flow impedance8, a decrease in the ratio of cerebral to placental Doppler indices (cerebroplacental ratio = CPR)9 or a decline in amniotic fluid volume10 as a sign of decreased renal perfusion. With fetal perception of hypoxemia additional vascular responses such as ‘brain sparing’ may be invoked to enhance local perfusion11, 12. Although often used interchangeably, redistribution (umbilical artery A/REDV), centralization (low CPR) and brain sparing appear to be different vascular responses. While the CPR is almost always abnormal in fetuses with overtly abnormal umbilical artery flow, brain sparing may develop independently of the umbilical artery waveform13-15. A number of changes in fetal biophysical variables are observed with chronic placental dysfunction. These may reflect abnormal maturation of brainstem reflexes, chronic hypoxemia, redistribution of cardiac output or a combination of these. Autonomic reflexes superimposed on the intrinsic cardiac activity determine fetal heart rate. These reflexes originate from the brainstem and may be modulated through incorporation of signals from higher centers, the reticular activating system and peripheral sensory inputs. Variations of the heart rate and episodic accelerations coupled to fetal movement are indicative of normal functioning of these connections. Conversely, lack of normal heart rate may represent hypoxia-mediated effects on the regulatory centers, decreased overall global fetal activity or delayed development of reactivity16-20. When the physiological responses to hypoxia become exhausted the fetus cannot adapt any further. Fetal decompensation may be associated with several findings. There is a decline in forward cardiac function21, associated with an increase in venous Doppler indices22-24. Deregulation of cardiovascular homeostasis may be seen (normalization of cerebral Doppler indices25, 26) and assessment of the arterial circulation becomes a less reliable index of compromise. Following a decline in global fetal activity, fetal breathing movement, body movements and tone may be lost18, 27. The development of overtly abnormal fetal heart rate patterns appears to be related to a significant worsening of cardiac dysfunction as a result of worsening hypoxemia28-30. It is apparent that the clinical presentation and disease spectrum of IUGR is variable reflecting the diverse pathophysiology. Several authors have suggested a sequence of deterioration that is evident on various antenatal surveillance tests. Abnormal umbilical artery flow, CPR and brain sparing are ‘early’ circulatory abnormalities10, 31, 32. Oligohydramnios, loss of fetal tone and or movement, abnormal venous flow and overt heart rate decelerations are typically ‘late’ changes10, 26, 27, 33. Yet such a rigid classification does not take into account individual variations in responses, the impact of gestational age, inconsistent relationships between circulatory and biophysical deterioration and the differential effects that sudden alterations in maternal status may have on fetal testing variables3, 27, 34-36. There is no uniform agreement on the relationship between computerized fetal heart rate changes and venous Doppler abnormalities33, 37. Studies by Hecher and Ferrazzi suggest that only 50% of fetuses with abnormal computerized CTG (cCTG) may develop abnormal venous indices10, 33. It is apparent, however, that arterial Doppler abnormalities identify a prodrome of fetal disease when the decline in biophysical variables is subtle and predominantly evident on computerized analysis. In the preterm fetus, deterioration of circulatory and biophysical status occurs in a close temporal relationship27. Because subtle Doppler findings are more common this relationship may not hold near term10, 38-40. In contrast, the loss in biophysical variables with metabolic deterioration is largely independent of the gestational age41. Absence of umbilical artery end-diastolic velocities indicates a significant villous abnormality. However, the relationship between placental pathology and fetal acidemia is inconsistent both at cordocentesis and birth28, 42-45. Brain sparing, elevation in thoracic and abdominal aortic pulsatility index (PI) and an abnormal CPR are all associated with a decrease in fetal pO2 and a median decrease of the pH of two standard deviations (2 SD)46-48. An elevation in precordial venous indices (inferior vena cava and ductus venosus) has to date provided the most consistent relationship with a significant decline in umbilical venous pH (∼ 4 SD) in IUGR fetuses48-50. Based on receiver–operating curve (ROC) statistics with a background acidemia rate of 41.5% and 52.8%, use of the inferior vena cava percentage reverse flow or the preload index provides sensitivities for the prediction of fetal acidemia ranging from 73.5% to 95% and specificities of 72.3% to 75%49, 50. And while the Doppler findings in each of the vascular beds correlate with fetal acid–base status, there is a wide variation of fetal pH with abnormal results (Figure 1). A diagrammatic representation of pH deviation from the gestational age mean (Δ pH) with abnormal test results in various antenatal tests. These include fetal heart rate (FHR) analysis using traditional non-stress testing (NST;—react, non-reactive)52 and the computerized cardiotocogram (cCTG; + acc, accelerations present; + dec, obvious decelerations present)29. Biophysical variables (AFV, amniotic fluid volume; FBM, fetal body movement; FGM, fetal gross movement)52. The same relationships are expressed for umbilical artery absent end-diastolic velocity (AEDV) and deviation of the arterial or venous Doppler index > 2 SD from the gestational age mean for the thoracic aorta (TAO), descending aorta (DAO), the middle cerebral artery (MCA), cerebroplacental ratio (CPR) and the ductus venosus (DV)28, 45-48. Similar observations have been made for fetal heart rate analysis, breathing, tone, gross body movement and amniotic fluid volume. While a reactive CTG even by criteria graded for gestational age virtually excludes hypoxemia, a non-reactive CTG is associated with a wide range of pH values41, 51, 52. The accuracy for the prediction of acidemia can be enhanced by computerized fetal heart rate analysis. All computerized variables such as short-term, long-term mean minute variation and episodic or periodic changes are related to a range of normal and abnormal fetal pH values29, 53, 54. In IUGR fetuses, a short-term variation of 3.5 ms due to prolonged episodes of low variation appears to be the best predictor of a cord artery pH < 7.20 at birth by ROC analysis with a background acidemia rate of 21%55. Though clearly abnormal in the non-laboring patient, lost in this analysis is the clinical relevance of a pH < 7.20 at delivery. Loss of fetal breathing movements is associated with a moderate decrease in pO2 and a wide range of pH in both cordocentesis samples and at birth. In contrast, the absence of fetal tone and gross body movement is almost always associated with acidemia41, 51, 52. Gestational age has a profound impact on all aspects of IUGR. The degree of placental pathology determines uteroplacental blood flow patterns, gestational age and spectrum of manifestation as well as the overall risk for adverse outcome. Mild placental dysfunction with minimal or no umbilical artery blood flow abnormality is more common among fetuses presenting with IUGR in the third trimester, and an abnormal CPR or brain sparing may be the only Doppler evidence of placental dysfunction. In contrast, umbilical artery A/REDV is rarely seen beyond 32–34 weeks since it is associated with early growth failure (second and early third trimester) marked fetal vascular and behavioral responses and early decompensation6, 10, 15, 40, 56. Abnormal venous flows are therefore almost exclusively reported in fetuses with markedly abnormal umbilical artery blood flow and early-onset IUGR. In the third trimester subtle Doppler findings in the arterial circulation are more characteristic10, 34, 39, 40, 57. There is a physiological change in the reference range for almost all Doppler indices with gestational age. To account for this, measurements need to be transformed to Z-scores for statistical analysis. An index deviation > 2 SD from the gestational age mean provides a widely accepted statistical cut-off for an abnormal result. The three arterial Doppler indices (S/D ratio, resistance index and PI) appear to vary little clinically58. It has not been resolved whether there are relevant differences between the numerous venous indices currently in use49. The establishment of fetal behavioral states and maturation of fetal heart rate control with coupling to fetal behavior is achieved at various gestational epochs. Delayed maturation of biophysical milestones is a feature of IUGR. The physiologic decline of the baseline heart rate and maturation of reactivity are delayed while short- and long-term variability is decreased59, 60. Similarly, development of behavioral states and integration of behavior patterns may be delayed61. These developmental abnormalities could influence the reliability of computerized, as well as traditional, fetal heart rate analysis. In addition to fetal manifestations, gestational age has a significant impact on short- and long-term outcomes. A/REDV is associated with an increased risk for stillbirth and premature delivery due to fetal decompensation36, 62. Fetal acidemia, poor transition to extrauterine life, condition of the neonate after delivery and degree of prematurity pose additional risks. IUGR fetuses are at higher risk for neonatal complications than their appropriately grown counterparts63-65. This risk is even higher for the preterm IUGR fetus and appears independent of Doppler status38, 66, 67. The summation of these factors is responsible for high morbidity, mortality and adverse neurodevelopmental outcome of the preterm IUGR fetus, particularly if delivered before 28 weeks66-69. These associations determine the relative balance of management decisions. Near-term delivery outweighs the risks of temporizing intervention and delivery is generally indicated. The risk for adverse neonatal outcomes is low and benefits of modified perinatal interventions are more likely to be reflected in long-term outcomes. In the preterm IUGR fetus intervention is triggered by balancing fetal and neonatal risks and its success is measured by impacts on short-term variables such as perinatal mortality and morbidity. The clinical spectrum of IUGR is wide, both in presentation and progression. Gestational age, differential fetal maturation, maternal condition, and therapeutic interventions (possibly) modulate the presentation and manifestation of fetal disease in various testing modalities. There is normal variation in biological parameters and their relationship with fetal acid–base status. As a result the prevalence of abnormal test results in a single testing modality and their relationship between acid–base status are inconsistent. Therefore, no single test provides well-validated cut-offs to accurately depict fetal status in IUGR. Two principal approaches may be taken to solve the problem of accurate fetal assessment. One approach to the diversity of the clinical spectrum is to increase the level of sophistication for testing modalities. Measurement of venous volume flow, incorporation of multiple oxygen-sensitive vascular beds and validation of venous flow abnormalities through measurements of the vessel diameter may be incorporated to enhance the sensitivity of Doppler surveillance23, 70-72. Similarly, computerized analysis of percentage of fetal activity, fractal fetal behavior, percentage of rapid eye movements and fetal movements during F2 activity states can be incorporated into the biophysical assessment17, 19. These techniques have yet to be validated and would substantially increase the technical complexity of routine fetal evaluation thus prohibiting large-scale application. An alternative approach is to combine antenatal testing modalities currently in use to improve diagnostic and predictive accuracy73. The logic and potential value of combining antenatal testing modalities that are independent is illustrated well by biophysical profile scoring. Each of the parameters of the BPS is independently altered by hypoxemia. Yet, the combination of the five components into a composite score performs better than each single parameter to predict fetal well-being. The BPS has a reproducible relationship with the fetal pH, perinatal morbidity and mortality from the mid-trimester onwards in IUGR fetuses41, 74, 75. As our understanding of fetal cardiovascular adaptation to IUGR has grown, Doppler surveillance has evolved to incorporate multiple vessels to enhance prediction of fetal acid–base status. A clinically validated risk stratification and management protocol comparable to the BPS has not yet been presented using multivessel Doppler. Given the variable manifestations of IUGR it is intuitive to incorporate an even broader range of practicable antenatal testing variables into the development of a surveillance protocol that will accurately assess fetal status. Fetal deterioration may be manifested by through abnormal behavior, central deregulation of cardiorespiratory function, alterations in vascular tone in oxygen-sensitive vessels and cardiovascular dysfunction. Each of the aforementioned fetal testing variables has an independent relationship with metabolic status. But only their combination evaluates the spectrum of cardiovascular and manifestations found in IUGR Doppler has been to our understanding of the relationship between cardiovascular and metabolic The alterations in Doppler velocities occur long before there are abnormalities of growth and acid–base As Doppler is a diagnostic that the of a during which an intervention could be the BPS the CTG this of information in the absence of overt of compromise. the of Doppler and BPS information in an assessment of the fetal condition and the of The same cannot be if assessment to venous Doppler for the of compromise in only of fetuses with early and an even near term10, 27, 34, Fetal deterioration of Doppler and biophysical variables in different Doppler changes compromise by while changes in amniotic fluid abnormal venous and decline in fetal breathing, tone and movement occur The integration of multivessel Doppler and BPS for of multiple patterns of placental insufficiency and fetal Doppler and BPS will both deterioration in severe early-onset In or IUGR with Doppler findings the BPS will fetal compromise. In hypoxemia or acidemia is likely to be in grown fetuses with placental dysfunction. If intervention is not the for surveillance can be on the of the condition and the of early Doppler This may include to a with level of perinatal for monitoring and of in of preterm The of fetal assessment to the condition can be achieved using widely The of combining assessment of fetal growth with of Doppler and biophysical parameters in the management of is not Doppler and are to between the but grown the growth fetus at risk for adverse umbilical artery Doppler with will need antenatal In the of antenatal and preterm delivery can be morbidity and mortality are outcomes that are currently used to impacts of perinatal effective fetal is currently and gestational age is a of the therapeutic of timing of delivery is our A analysis by the that delivery by to 2 weeks can an of in the fetus overall perinatal Therefore, temporizing intervention appears between and each in may neonatal mortality by However, the increased stillbirth rate when delivery is triggered by overt abnormalities of fetal testing such as decelerations the of better to fetal compromise. Several authors have the of assessment. and used ROC analysis and that the combination of artery Doppler and traditional CTG prediction of fetal and low and the use of or stress and reported assessment of fetal among fetuses using the modified biophysical score in combination with umbilical and middle cerebral artery Doppler Doppler abnormalities used to determine the of fetal testing and to the of fetal It has been suggested that standard CTG with computerized analysis the prediction of fetal and mortality in IUGR A and a fetal components used in the evaluation of the neonate with a combination of Doppler and biophysical and expressed through the uterine artery resistance index and the artery Doppler ratio, by BPS reflexes by and the heart rate with the The fetal score comparable to the BPS in the prediction of fetal and poor transition to extrauterine life, but in IUGR fetuses with acidemia at to that the combination of the BPS with multivessel arterial and venous Doppler is better in the prediction of outcomes than modality In this prediction of perinatal acidemia at birth and neonatal While the of associations provides insight into the relationships between fetal status and the most evidence is provided by clinical management and suggested that IUGR is by and umbilical artery Doppler analysis, of BPS can management by accurately the risk for acidemia and stillbirth of gestational age and Doppler Similarly, and that the combination of umbilical artery Doppler and biophysical profile In a randomized management fetuses with with BPS and delivered for worsening maternal status, BPS < or to fetuses with end-diastolic there no differences in pH and low at birth. there no perinatal in any of the fetal surveillance using multivessel Doppler and BPS is likely to enhance surveillance by early of deterioration and in the preterm IUGR of long-term outcomes is and associations are predominantly on of management has in a 50% of cerebral in the In relationships between the score and and have been Doppler findings such brain sparing, absent and reversed umbilical artery end-diastolic velocity have been associated with a variety of neurodevelopmental It to that of fetal testing to a population of IUGR fetuses is likely to perinatal of long-term outcome. CTG and BPS are widely techniques that have evolved the It is that these monitoring a of information on fetal and long-term risks. It is that their combination a better understanding of fetal in IUGR. of the such as and and local of are less is is should be and whether an intervention will improve long-term outcome. In addition need more information on to when test results are in with each other. The combination a randomized of management Doppler and biophysical information should be accurately in a in any randomized the impacts of interventions in IUGR. Irrespective of interventions and such an approach analysis of multiple variables for their predictive accuracy of short- and long-term outcomes. This will allow of fetal testing to the management in fetal growth under any