© 2012 by the American Institute of Ultrasound in Medicine
More than a decade ago, Mari et al1,2 achieved a major breakthrough in the treatment of Rh-sensitized fetuses with their pioneering work that showed a correlation between the Doppler middle cerebral artery peak systolic velocity (PSV) and fetal hemoglobin levels. This technique has virtually eliminated the need for invasive procedures such as amniocentesis and cordocentesis that have been used for diagnosis of fetal anemia with their inherent complications. Since then, the middle cerebral artery PSV has been the standard of care for treatment of anemic fetuses. Doppler studies have also been used in neonates with different cerebral conditions (eg, intraventricular hemorrhage, brain lesions, and hydrocephalus).3–6 However, this approach has not been attempted in neonates suspected to have blood volume disorders such as anemia and polycythemia. If a correlation between neonatal hemoglobin levels and the middle cerebral artery PSV is found, it may be similarly applied for rapid, noninvasive bedside diagnosis of acute life-threatening conditions in neonates until the standard blood tests can be performed.
The aims of this study were therefore to determine whether a correlation exists between the neonatal middle cerebral artery PSV and hemoglobin levels and to assess the possibility of implementing this indicator for rapid, noninvasive diagnosis of blood volume disorders in neonates.Materials and Methods
This prospective study included 151 healthy neonates, weight appropriate for gestational age, born at our medical center during a 6-month period. All neonates were delivered at 37 weeks’ gestation or later with Apgar scores of 7 or higher at 5 minutes. We excluded all neonates with malformations, intrauterine growth restriction, perinatal asphyxia, and infections. The local Research Ethics Institutional Review Committee approved the study, and informed consent was obtained from the mother of each neonate enrolled in the study. Medical examinations by a senior pediatrician confirmed that all neonates enrolled in the study were healthy without dysmorphic features. The neonates were born by spontaneous vaginal delivery or cesarean delivery. All neonates were prospectively studied on the second day of life (between 24 and 36 hours after delivery). Anemia was defined as a hemoglobin level of 13.5 g/dL or less or a hematocrit value 45% or less, and polycythemia was defined as a hemoglobin level greater than 22 g/dL or a hematocrit value greater than 65%.7
Doppler examinations of the middle cerebral artery were performed with a Voluson 730 ultrasound system (GE Healthcare,
) and a
convex transducer (4–8 MHz) in a quiet room. The neonates were sleeping in a
crib without gross body or limb movements and were breathing quietly. The
examinations were performed by using an axial plane on the temporal bone anterior
to the external auditory canal and superior to the zygomatic process,
identifying the middle cerebral artery. Measurements were obtained just distal
to the middle cerebral artery origin from the internal carotid artery. The
angle of insonation was close to 0°, thus obviating the need for angle
correction (Figure 1).
The sample gate was 3 to 4 mm. The total examination time was 1 to 3 minutes.
Five Doppler waves were recorded, and the highest PSV waveform was used for
An analysis of variance was performed to evaluate the different variables in the 3 groups studied (normal, anemic, and polycythemic neonates). Multiple comparison analyses were performed as well to determine whether the variable means were statistically different from each other. A regression analysis was conducted to test correlations between hemoglobin levels and middle cerebral artery PSVs in the whole groups. P < .05 was considered significant.
The study population included 122 normocythemic, 24 anemic, and 5 polycythemic neonates. The mean gestational age ± SD of the neonates at delivery was 39 ± 1.5 weeks, with a median Apgar score of 10 at 5 minutes and a mean birth weight of 3290 ± 446 g.
Table 1 presents the hemoglobin, hematocrit, and PSV values of the 3 groups. There were significant differences in the hemoglobin, hematocrit, and PSV values between the normocythemic neonates and the anemic and polycythemic neonates (P < .001). Of the 24 anemic neonates, 20 (83%) had a middle cerebral artery PSV that was higher than the 95% confidence interval (CI) for normocythemic neonates, and all 5 polycythemic neonates had a PSV that was lower than the 95% CI for normocythemic neonates.
In Figure 2, the means and 95% CIs of the middle cerebral artery PSV values in the 3 groups (anemic, normocythemic, and polycythemic) are shown. Although there are overlapping values, the means of the 3 groups are significantly different and can be easily distinguished from each other (P < .01). Figure 3 depicts the middle cerebral artery PSV according to different hemoglobin levels (with the means and 95 percent CIs). A clear decrease in the PSV is evident with increasing hemoglobin levels (P < .01).
In Figure 4, the middle cerebral artery PSV of the 3 groups combined is depicted with a third-order polynomial fit regression line. Although there are overlapping values, the trend is clear (especially at the extremes of the hemoglobin levels) that the lower the hemoglobin concentration, the higher the PSV and vice versa. A significant correlation between the PSV and hemoglobin levels was found (P < .01). It is interesting to note that a plateau exists at hemoglobin levels considered to be within the normal range (±2 SDs) for neonates (at ≈13–22 g/dL), but below or above these limits, there are acute changes in the PSV.
In Figure 5, we show a vector plot of several anemic neonates who underwent partial exchange transfusion. The hemoglobin level and middle cerebral artery PSV were obtained at the bedside before the blood transfusion and 1 hour after the transfusion. The plot shows the trend of changes in the PSV, and the lines span from the starting to ending hemoglobin levels. Once more, it is clear that an increasing hemoglobin level caused an immediate decrease in the PSV. These fetuses with initial hemoglobin levels of 7.8 to 11.9 g/dL had PSVs of 51 to 144 cm/s, which rapidly decreased to approximately 32 to 80 cm/s when the hemoglobin levels increased to the normal range (>13 g/dL).
This study shows that there is a significant correlation between hemoglobin levels and the middle cerebral artery PSV in neonates. Although overlapping measurements in the normal range of hemoglobin levels exist, the more severe degrees of anemia and polycythemia can be readily diagnosed by examining the middle cerebral artery PSV. Similarly to the well-established technique used in fetuses, this procedure can also be suitable in neonates for prompt diagnosis of life-threatening blood volume disorders. Obviously, we do not imply that this method can replace the traditional direct blood examination. However, it may be used as an ancillary, rapid means of noninvasively estimating the degree of anemia or polycythemia in neonates suspected to have blood volume disorders, thus allowing prompt action. There are several neonatal conditions in which middle cerebral artery PSV can be rapidly used for diagnosis of anemia and polycythemia. These include anticipated deliveries of twins affected by twin-twin transfusion syndrome, neonates affected by Rh and Kell isoimmunization, thalassemia, parvovirus B19 infection, and hydrops. In addition, acute intrapartum events such as intracranial hemorrhage, a large cephalhematoma, and other hemorrhages associated with traumatic instrumental delivery with loss of a substantial amount of blood or a sudden decrease in blood volume can be immediately recognized at the bedside when clinical suspicion dictates. Even critically ill neonates for whom venous access is difficult (eg, hydropic neonates) can have a prompt diagnosis until blood tests can be safely performed. The appealing aspect of this technique is that it can be easily studied and mastered even by novice users in a very short period.
The underlying pathophysiologic mechanism of increased cardiac output and decreased blood viscosity in anemic fetuses is valid also for neonates, as shown in this study. Anemia causes an increase in the cardiac stroke volume, heart rate, and peripheral resistance and decreased blood viscosity, leading to an increase in cerebral blood flow to maintain adequate oxygen transport to the brain.8,9 Neonates, similarly to fetuses, obey the same physical laws of flow velocities in blood vessels.
Polycythemia, on the other hand, occurs in 2% to 5% of term neonates,10 usually as a compensatory mechanism in intrauterine hypoxia or uncontrolled diabetic pregnancies with macrosomic neonates or as a result of delayed cord clamping.11 This condition may lead to hyperviscosity of the blood with altered rheologic properties and flow disturbances, which can result in impaired perfusion to multiple organs. This situation can cause neurologic, cardiorespiratory, gastrointestinal, and renal abnormalities.12–15 Although these symptoms are usually transient, prompt diagnosis and treatment may be life saving with reversal of the potential damage.16,17
We found that for the established normal range of hemoglobin levels, the middle cerebral artery PSV has a plateau, whereas in anemia and polycythemia, the PSV changes rapidly (increasing and decreasing, respectively). The correlation between hemoglobin levels and the PSV becomes more pronounced as the severity of anemia or polycythemia increases (Figure 2). This factor may be due to the rheologic properties of the blood in neonates. Flow remains almost constant for a wide range of hemoglobin levels but rapidly changes as the hemoglobin levels decrease or increase beyond certain limits. It is interesting to note that the middle cerebral artery PSVs of the term neonates in this study (Table 1) were very similar to those reported by Mari et al2 in term fetuses, and anemic fetuses had PSVs similar to those of anemic neonates.
As to the question of whether this technique can be implemented in clinical practice, we have shown several neonates who underwent partial exchange transfusion because of anemia and were studied with the Doppler middle cerebral artery PSV before and after the transfusion (Figure 4). It is evident that normalizing the hemoglobin level rapidly corrects the PSV. We think that in polycythemic neonates, the contrary occurs as well.
This study had some limitations. We studied only term neonates 24 to 36 hours after delivery, examining our hypothesis that the Doppler middle cerebral artery PSV can be helpful in managing neonatal emergencies occurring in the first days after delivery (eg, intracerebral bleeding due to traumatic delivery). Because it has been reported that the PSV progressively changes during the first month of life,18 the flow velocities may be different later, and caution should be used in interpreting hemoglobin levels as a function of the middle cerebral artery PSV in older neonates.
Although it is appealing to also use this technique in premature neonates to diagnose acute anemia caused by massive hemorrhage, a caveat should be addressed in this group as well. The situation may be different in premature neonates in whom the proportion of hemoglobin F is different, and there may be different rheologic properties of the blood due to a different elasticity or size of the red blood cells. This issue should be further studied in the future. An additional factor that was not controlled for but may potentially alter the middle cerebral artery PSV is the presence or absence of a patent ductus arteriosus. However, the impact of the ductus on blood flow to the brain has been reported to be minimal19; therefore, we think that this factor may have only a marginal effect on middle cerebral artery PSV measurements in anemic and polycythemic neonates.
In conclusion, Doppler measurement of the middle cerebral artery PSV appears to be helpful for estimating the hemoglobin concentration in neonates and can be used as a screening tool for diagnosing neonatal anemia and polycythemia. This technique may allow a rapid, noninvasive determination of the neonatal hemoglobin level, dictating the urgency of treatment.