Changes in regional fetal cerebral blood flow perfusion in relation to hemodynamic deterioration in severe intrauterine growth restricted (IUGR) fetuses
Background
The aim was to evaluate changes in blood perfusion in three different brain regions, in appropriate for gestational age (AGA) fetuses, measured with fractional moving blood volume (FMBV) in relation to the hemodynamic adaptive stages of fetal deterioration, in intrauterine growth restricted (IUGR) fetuses.
Intrauterine growth restriction (IUGR) associated with placental insufficiency may be complicated with adverse neurological outcomes in up to 50-60% of cases. While overt brain lesions such as leukomalacia and/or cerebral palsy affect a relatively small number of IUGR infants, a substantial proportion will present a wide spectrum of subtle disturbances. Long-term studies demonstrate that neurodevelopmental dysfunction in IUGR mainly involves general cognitive competence, suggesting dysfunction in the frontal lobe networking, limbic system and hippocampus, as well as changes in the morphology of neural structures such as the retinal optical nerve.
The presence of neurological damage originating in different brain areas can be associated with an impaired blood supply. The blood flow centralisation process, also known as the brain sparing effect, is an essential adaptive response to preserve brain oxygen supply in the presence of chronic hypoxia. This process is clinically identified by a reduced Doppler pulsatility index (PI) in the middle cerebral artery (MCA). This concept implies that the magnitude of the increase in blood perfusion is similar in the entire cerebral circulation. However, recent research has suggested the presence of regional brain redistribution of the blood flow in relation to the intensity and duration of the hypoxic insult. The potential existence of regional brain redistribution in human fetuses with severe growth restriction, as well as its progression during fetal deterioration, have not been evaluated.
Blood flow perfusion can be reliably estimated using power Doppler ultrasound (PDU) and fractional moving blood volume (FMBV) measurement, which has acceptable reproducibility and shows a high correlation with real blood perfusion changes.
Method
B 56 IUGR and 56 appropriate for gestational age (AGA) fetuses matched by gestational and maternal age were evaluated. Data from 35 AGA fetuses included in our previous methodological report were included in this study. Gestational age at the time of the study was (median) 29 (range, 26 - 32) weeks of gestation, and maternal age was (median) 32 years (range 20 - 39 years). The study was approved by the hospital’s ethics committee and written consent from all participants was collected.
IUGR was defined as an ultrasound estimated fetal weight below the 10th centile and an abnormal PI (mean>2SD) in the umbilical artery (UA), both for gestational age. Based on the abnormal UA, middle cerebral artery, (MCA), and ductus venosus, (DV), waveforms, the fetuses were divided into the following groups:
group 1: abnormal UA pulsatility index (PI), (mean >2SD, n=13)
group 2: abnormal UA and MCA PI (mean <2SD n=15)
group 3: abnormal UA, MCA, and ductus venosus DV-PI (mean >2SD) but present atrial (a wave) flow (n=16)
group 4: absent or reversed DV atrial flow (n=12).
All studies were performed with Siemens Antares ultrasound equipment using a 6-2 MHz linear curved array in the absence of fetal respiratory and corporal movements, and with the mother in voluntary suspended respiration. The mechanical and thermal indices were always maintained below 1. Standard Doppler evaluation of the UA was performed in a free loop of the umbilical cord, MCA at the level of its origin from the circle of Willis, and ductus venosus after its emergence from the portal sinus. Doppler recordings were obtained, maintaining the angle of insonation as close as possible to 0°. We used manual correction when the angle was not 0° and if the angle was more than 30° we excluded that recording. A high pass wall filter (70 Hz) was used to eliminate low frequency signals. At least three consecutive good quality waveforms from each vessel were automatically measured. The process was repeated 3 times and the mean was used as the representative value of that fetus.
For blood perfusion estimation, PDU was applied using the following fixed settings: standard grey scale image for obstetrics, medium persistence, high sensitivity, normal image display, pulse repetition frequency 610 Hz, medium wall filter, and gain level just above the presence of noise. The following anatomical planes of the fetal brain were included in the PDU colour box: midsagittal for the anterior and complete midsagittal FMBV estimations, parasagittal for basal ganglia, and transverse at the level of the posterior fossa for cerebellar FMBV. For midsagittal and parasagittal views, special care was taken to insonate through the sagittal suture in order to avoid the bone barrier and to have an uncluttered PDU recording.
A sequence of at least 10 consecutive good quality images in the absence of flash artifacts from each region was stored. These images covered at least the time of two complete cardiac cycles, thus averaging the values obtained during systole and diastole. PDU images for the described anatomical planes were obtained in all fetuses, disregarding the fetal position. Only examinations during the first visit to the IUGR clinic were included in the analysis. Images were then transferred to a PACS system in a TIFF format and analysed in purpose-designed software. The regions of interest (ROIs) were delimited off-line as follows: anterior cerebral, in an anterior mid-sagittal view of the fetal head, delimited anteriorly by the internal wall of the skull, posteriorly by an imaginary line drawn at 90° at the level of the origin of the anterior cerebral artery, and parallel to an imaginary line in the front of the face, and inferiorly by the base of the skull (Figure 1a); complete midcerebral, in a midsagittal plane delimited by the complete internal wall of the fetal skull (Figure 1b); basal ganglia in a midparasagittal view, delimited by the head, body and tail of the caudate nucleus and inferiorly by the lenticular nucleus (Figure 1c), and the cerebellum delimited anteriorly by the base of the cerebellar hemispheres and posteriorly by the fetal skull (Figure 1d). Pixels containing colour PDU ultrasound data within the ROIs were used for FMBV calculation. The mean FMBV from all 10 images was considered as the representative value for each fetus. Differences between AGA and IUGR fetuses were evaluated using Student t-test, and differences within different IUGR groups were evaluated by analysis of variance and Tukey’s post hoc-test. A p-value of <0.05 was considered significant.
Table 1:
Clinical characteristics of the studied population at birth
| |
|
|
|
|
|
| |
|
Group 1 (n=13) |
Group 2 (n=15) |
Group 3 (n=16) |
Group 4 (N=12)_ |
|
GA at delivery(weeks) |
39 (4) |
35 (3) |
33 (2) |
30 (3) |
28 (2) |
|
Cesarean section (%) |
23 |
89 |
96 |
95 |
100 |
| Birth weight (g) |
3082 (911) |
1086 (280) |
1028 (279) |
962 (297) |
702 (251) |
| 5-min Apgar<7 |
0 |
0 |
0 |
3 (18.7%) |
5 (41.6%) |
| Arterial cord
pH<7.20 |
0 |
1 (7%) |
1 (6%) |
3 (18.7%) |
7 (58.3%) |
| Days in NICU |
0 |
43 (16) |
41 (24) |
37 (16) |
41 (23) |
|
In-utero mortality (%) |
0 |
1 (8%) |
0 |
1 (6%) |
2 (16%) |
|
Neonatal mortality (%) |
0 |
0 |
1 (6%) |
2 (12%) |
5 (41%) |
|
Perinatal mortality (%) |
0 |
1 (8%) |
1 (6%) |
3 (19%) |
6 (50%) |
NICU neonatal intensive care unit, APO adverse perinatal outcome (neonatal parenchymalbrain damage, intraventricular hemorrhage, severe pulmonary distress syndrome, necrotising enterocolitis, renal failure or perinatal death), IUGR S1, UA-PI>2SD; S2, MCA-PI<-2SD; S3, DV-PI>2SD; S4, DV-PI>5SD.
Perinatal mortality defined as intrauterine plus neonatal (first 28 days of life) death * P<0.05 compared with term AGA, †P<0.05 compared with preterm AGA.
Table 1 shows the perinatal results of the study population. Gestational age and birth weight showed a decreasing trend from groups 1 to 4. Almost all neonates were delivered by caesarean section, either because they reached a gestational age with good prognosis or because they showed signs of fetal deterioration. Groups 3 and 4 showed a higher prevalence of reduced Apgar scores at 5 minutes and low pH values in the umbilical cord. The length of stay in the neonatal intensive care unit was similar in all groups; only surviving fetuses were included in this analysis. Perinatal mortality and the presence of adverse perinatal outcomes were more frequent in group 4 but no statistically significant differences were observed in comparison with the remaining groups, most probably because of the sample size.
Table 2: Fractional moving blood volume (FMBV) values (mean, SD) obtained in the studied cerebral regions of interest in intrauterine growth restricted fetuses (IUGR) at different stages of hemodynamic deterioration.
| Controls( n+56) |
17.1 (4.0) |
18.6 (4.7) |
14.0 (3.4) |
6.4 (2.8) |
| All IUGR (n+56) |
28.5 (7.2) ** |
26.3 (8.2) ** |
24.7 (6.9) ** |
11.4 (4.3) ** |
| Group 1 (n=13) |
33.2 (7.1)*† |
21.6 (4.9)* |
28.3 (6.8)* |
11.8 (4.3)* |
| Group 2 (n=15) |
27.6 (6.8)* |
25.0 (6.4)*Ψ |
21.9 (5.8)* |
11.3 (3.5) * |
| Group 3 (n=16) |
28.3 (6.9)* |
29.2 (7.7)*‡ |
25.2 (7.1)* |
12.7 (4.6)* |
-
Group 1 - abnormal umbilical artery (UA) pulsatility index (PI) (mean >2SD)
-
Group 2 - abnormal umbilical and middle cerebral artery (MCA) PI (mean <2SD),
-
Group 3 - abnormal UA, MCA, and ductus venosus (DV) PI (mean >2SD) but present atrial flow and
-
Group 4 - absent or reversed DV atrial flow.
** Student T test IUGR vs. controls: p<0.05; ANOVA (p<0.05),
* vs. controls;
† vs. groups 2, 3, 4;
¥ vs. groups 1, 2, 3;
‡ vs. groups 1, 2;
Ψ vs. group 1.
FMBV values obtained from the studied regions are shown in Table 2. In all ROIs, FMBV values were significantly increased in IUGR fetuses compared with those in AGA fetuses. In relation to fetal hemodynamic deterioration, there was an acute increment in the anterior sagittal FMBV in group 1, followed by a constant significant reduction toward group 4 (F= 3.25, p=0.027). In contrast, basal ganglia FMBV showed a constant increment (F=11.61, p<0.001) as the fetus became hemodynamically more affected. Complete sagittal FMBV followed the same trend as the anterior sagittal FMBV, but no statistically significant differences within the IUGR stages were noted, (Table 2). Cerebellar FMBV showed a significant increment in IUGR group 1 with no significant changes in the remaining groups. Net FMBV values from the cerebellar ROI cannot be compared with FMBV values from other ROIs as the bone shadow significantly reduces backscattered echoes.
Figure 1:
Increment (in percentage) of FMBV values in the different IUGR hemodynamic groups as compared with control values from all the studied regions ofinterest.
Group 1 - abnormal umbilical artery (UA) pulsatility index (PI) (mean >2SD)
Group 2 - abnormal umbilical and middle cerebral artery (MCA) PI (mean <2SD)
Group 3 - abnormal UA, MCA, and ductus venosus (DV) PI (mean >2SD) but present atrial flow and
Group 4 - absent or reversed DV atrial flow.
Figure 1 shows the percentage of increment of FMBV in the studied ROIs as compared with normal FMBV values. Whereas in IUGR group 1, complete sagittal, anterior sagittal and cerebellar FMBV showed the highest increment, basal ganglia FMBV presented a mild-to-moderate increment. In further IUGR stages, this increment in the basal ganglia and cerebellar FMBV was more evident, whereas the anterior and complete sagittal FMBV tended to decrease.
Figure 2
Power Doppler images from the studied regions of interest (ROI) where fractional moving blood volume (FMBV) was estimated;
(a) anterior midsagittal: ROI is delineated posteriorly, by an imaginary line emerging at 90 degrees from the beginning of the anterior cerebral artery (ACA), and parallel to an imaginary line in the front of the face; the internal wall of the skull completed the anterior, superior and inferior limits
(b) complete midsagittal: ROI is delimited by the complete internal wall of the skull with a clear image of the ACA, pericallosal artery (PCLA), sagittal sinus (SS) and Galen vein (GV)
(c) basal ganglia obtained in a midparasagittal view of the fetal brain: ROI is delimited by the head and body of the caudate nucleus (CN) which forms the floor of the lateral horn of the cerebral ventricles and
(d) cerebellar region delimited anteriorly by the beginning of the two posterior cerebral arteries (PCA), and posteriorly by the internal wall of the skull.
Discussion
This study provides evidence that brain blood perfusion in fetal growth restriction follows an internal regional redistribution, which changes substantially with progression of hypoxic fetal deterioration. The anterior cerebral area showed a two-fold increase in FMBV values in the first IUGR group but then declined steadily from group 2, reaching half the initial increment in group 4. In contrast, perfusion in the basal ganglia showed a progressive increment, from an initial 16% increment to nearly 100% in later groups of fetal deterioration. Although variation was less pronounced, FMBV in the cerebellum followed a similar progression to that of the basal ganglia.
The initial preferential increment in blood supply to the anterior frontal lobe can be associated with protection of general cognitive functions such as impulse control, language, memory, problem solving, and socialisation. The further increment in the basal ganglia and cerebellum can be related to protection of motor functions such as movement and postural control. Both areas also present a clear feedback with the cortex through the thalamus. Whereas control of basic functions such as blood pressure, breathing and cardiac regulation are located in the brainstem, there is a rich connection with the basal ganglia and thalamus. The existence of regional brain blood flow redistribution suggests a hierarchical order in the protection of brain functions according to the severity of the hypoxic insult.
Our findings are in line with recent studies in IUGR fetuses suggesting that the anterior cerebral artery shows Doppler signs of vasodilation before these are observed in the MCA 1. Indeed, in this study, the observation of an abnormally reduced PI in the MCA was concurrent with the onset of the progressive reduction in blood perfusion to the anterior area. These data might suggest that, contrary to current beliefs, an abnormal MCA might actually indicate the starting point after which the hemodynamic protection of the anterior area starts to decline. If confirmed, these findings may have important implications and deserve long-term follow-up studies.
The results of the present study and those of previous animal studies suggest that the protective effect of increased blood perfusion may initially be more intense in areas controlling higher functions but from early stages of fetal hypoxic deterioration, may progressively shift to brain regions controlling basic functions for survival.
This study has several potential limitations. The first is that FMBV is an indirect estimate of tissue perfusion. However, the technique has shown a strong correlation with gold standards in experimental conditions16 and we have previously shown that brain FMBV can be reliably and consistently be estimated in human fetuses. Our results are consistent with those of Dubiel et al. who also reported that IUGR foetuses have increased brain power Doppler signals, although these authors did not evaluate FMBV or the existence of regional variations. Analysis of 10 consecutive images involved at least 2-3 cardiac cycles. Although our results showed a consistent pattern, they may not completely represent physiological changes. The same argument can be applied to other Doppler modalities. Pulsed Doppler evaluation of different vascular territories under the same physiological influence included a minimum of 3 complete waveforms. Increasing the number of analysed images would have increased the time spent without producing real changes in the final results. The guidelines to evaluate FMBV have previously been reported.
Secondly, we focused this first study in only four areas, as these areas can be clearly delineated and show acceptable inter-observer variability. However, examination of further cerebral regions, particularly the brainstem, would be needed to complement these observations. Standardisation of further FMBV brain sections to include the brainstem, or study of specific brain structures, is now underway.
The clinical significance of the observations reported in the present study remains to be established by prospective studies including longterm postnatal neurological follow-up. However, correlation of prenatal hemodynamic findings and long-term outcome is lacking for most Doppler signs currently used in clinical practice. In the absence of this information, most current clinical protocols for fetal growth restriction are based on the assumptions that the onset of a brain sparing effect is indicated by a reduced PI in the MCA Doppler, and that this sign represents the establishment of a protective hemodynamic response in the entire fetal brain. The results of this study failed to support these beliefs, and indicate the need for future prospective studies to establish the correlation of hemodynamic cerebral changes with the onset of brain damage in fetal growth restriction.
Conclusion
The hemodynamic adaptive changes in IUGR fetuses result in significant regional variations in brain blood perfusion. The anterior area seems to be the first to be protected, but as the fetus becomes more affected, the regional blood redistribution process increases continuously towards the basal ganglia, with no further changes in the anterior or cerebellar regions.
This information is not meant to replace the advice of any physician or qualified health professional. The information provided by Cerebra is for information purposes only and is not a substitute for medical advice or treatment for any medical condition. You should promptly seek professional medical assistance if you have concerns regarding any health issue.